{"id":68851,"date":"2026-01-30T19:49:01","date_gmt":"2026-01-30T11:49:01","guid":{"rendered":"https:\/\/www.wsisp.com\/helps\/68851.html"},"modified":"2026-01-30T19:49:01","modified_gmt":"2026-01-30T11:49:01","slug":"rag%e6%a3%80%e7%b4%a2%e5%a2%9e%e5%bc%ba%e4%bc%98%e5%8c%96%e6%8c%87%e5%8d%97","status":"publish","type":"post","link":"https:\/\/www.wsisp.com\/helps\/68851.html","title":{"rendered":"RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357"},"content":{"rendered":"<\/p>\n<h4>\u76ee\u5f55<\/h4>\n<ul>\n<li>RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357<\/li>\n<li>\n<ul>\n<li>\u4e00\u3001\u4e09\u4ee3RAG\u67b6\u6784\u5bf9\u6bd4<\/li>\n<li>\n<ul>\n<li>1.1 \u67b6\u6784\u6f14\u8fdb\u65f6\u95f4\u7ebf<\/li>\n<li>1.2 \u4e09\u4ee3\u67b6\u6784\u8be6\u7ec6\u5bf9\u6bd4<\/li>\n<li>1.3 Naive RAG &#8211; \u57fa\u7840\u67b6\u6784<\/li>\n<li>\n<ul>\n<li>Python\u5b9e\u73b0\u793a\u4f8b<\/li>\n<\/ul>\n<\/li>\n<li>1.4 Advanced RAG &#8211; \u589e\u5f3a\u67b6\u6784<\/li>\n<li>1.5 Modular RAG &#8211; \u6a21\u5757\u5316\u67b6\u6784<\/li>\n<li>1.6 RAG \u5168\u94fe\u8def\u589e\u5f3a\u8def\u7ebf<\/li>\n<\/ul>\n<\/li>\n<li>\u4e8c\u3001\u68c0\u7d22\u524d\u589e\u5f3a<\/li>\n<li>\n<ul>\n<li>2.1 HyDE (Hypothetical Document Embeddings)<\/li>\n<li>2.2 Query2Doc<\/li>\n<li>2.3 \u67e5\u8be2\u91cd\u5199 (Query Rewriting)<\/li>\n<li>2.4 \u67e5\u8be2\u5206\u89e3 (Query Decomposition)<\/li>\n<li>2.5 Take Step Back<\/li>\n<\/ul>\n<\/li>\n<li>\u4e09\u3001\u68c0\u7d22\u589e\u5f3a<\/li>\n<li>\n<ul>\n<li>3.1 \u6838\u5fc3\u6982\u5ff5<\/li>\n<li>3.2 \u6df7\u5408\u68c0\u7d22 (Hybrid Search)<\/li>\n<li>\n<ul>\n<li>BM25\u7b97\u6cd5\u539f\u7406<\/li>\n<li>\u5411\u91cf\u68c0\u7d22\u539f\u7406<\/li>\n<li>RRF\u878d\u5408\u7b97\u6cd5<\/li>\n<\/ul>\n<\/li>\n<li>3.3 \u53e5\u5b50\u7a97\u53e3\u68c0\u7d22 (Sentence Window Retriever)<\/li>\n<li>3.4 \u7236\u5b50\u6587\u6863\u68c0\u7d22 (Parent Document Retriever)<\/li>\n<li>3.5 \u6458\u8981\u68c0\u7d22 (Summary Retriever)<\/li>\n<li>3.6 \u5047\u8bbe\u6027\u95ee\u9898\u68c0\u7d22 (Hypothetical Questions Retriever)<\/li>\n<li>3.7 \u591a\u7d22\u5f15\u878d\u5408\u7b56\u7565<\/li>\n<\/ul>\n<\/li>\n<li>\u56db\u3001\u68c0\u7d22\u540e\u589e\u5f3a<\/li>\n<li>\n<ul>\n<li>4.1 \u91cd\u6392\u5e8f (Re-ranking) \u539f\u7406<\/li>\n<li>4.2 \u4e3b\u6d41Re-ranker\u5bf9\u6bd4<\/li>\n<li>4.3 Re-ranking\u5b9e\u73b0<\/li>\n<li>4.4 \u4e0a\u4e0b\u6587\u538b\u7f29<\/li>\n<\/ul>\n<\/li>\n<li>\u4e94\u3001\u5b8c\u6574\u751f\u4ea7\u7ea7\u793a\u4f8b<\/li>\n<li>\n<ul>\n<li>5.1 LangChain\u751f\u4ea7\u7ea7\u5b9e\u73b0<\/li>\n<li>5.2 LlamaIndex\u751f\u4ea7\u7ea7\u5b9e\u73b0<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h2>RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357<\/h2>\n<p>\u6838\u5fc3\u6982\u5ff5 RAG (Retrieval-Augmented Generation) \u662f\u4e00\u79cd\u9769\u547d\u6027\u7684AI\u67b6\u6784&#xff0c;\u5de7\u5999\u5730\u5c06\u68c0\u7d22\u7cfb\u7edf\u4e0e\u751f\u6210\u5f0f\u5927\u8bed\u8a00\u6a21\u578b\u7ed3\u5408\u3002\u5b83\u50cf\u7ed9\u5927\u6a21\u578b\u914d\u5907\u4e86\u4e00\u4e2a&#034;\u5916\u6302\u77e5\u8bc6\u5e93&#034;&#xff0c;\u8ba9AI\u80fd\u591f&#xff1a;<\/p>\n<ul>\n<li>\u4ece\u5916\u90e8\u77e5\u8bc6\u5e93\u5b9e\u65f6\u83b7\u53d6\u4fe1\u606f<\/li>\n<li>\u57fa\u4e8e\u4e8b\u5b9e\u751f\u6210\u7b54\u6848&#xff0c;\u5927\u5e45\u51cf\u5c11\u5e7b\u89c9<\/li>\n<li>\u77e5\u8bc6\u5e93\u53ef\u968f\u65f6\u66f4\u65b0&#xff0c;\u65e0\u9700\u91cd\u65b0\u8bad\u7ec3\u6a21\u578b<\/li>\n<li>\u63d0\u4f9b\u7b54\u6848\u6765\u6e90&#xff0c;\u589e\u5f3a\u53ef\u89e3\u91ca\u6027<\/li>\n<\/ul>\n<h3>\u4e00\u3001\u4e09\u4ee3RAG\u67b6\u6784\u5bf9\u6bd4<\/h3>\n<p>RAG\u6280\u672f\u7ecf\u5386\u4e86\u4e09\u4ee3\u6f14\u8fdb&#xff0c;\u6bcf\u4e00\u4ee3\u90fd\u89e3\u51b3\u4e86\u524d\u4e00\u4ee3\u7684\u75db\u70b9\u3002<\/p>\n<h4>1.1 \u67b6\u6784\u6f14\u8fdb\u65f6\u95f4\u7ebf<\/h4>\n<p>  #mermaid-svg-cEUNC1ekp3T38YEU{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-cEUNC1ekp3T38YEU .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-cEUNC1ekp3T38YEU .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-cEUNC1ekp3T38YEU .error-icon{fill:#552222;}#mermaid-svg-cEUNC1ekp3T38YEU .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-cEUNC1ekp3T38YEU .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-cEUNC1ekp3T38YEU .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-cEUNC1ekp3T38YEU .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-cEUNC1ekp3T38YEU .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-cEUNC1ekp3T38YEU .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-cEUNC1ekp3T38YEU .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-cEUNC1ekp3T38YEU .marker{fill:#333333;stroke:#333333;}#mermaid-svg-cEUNC1ekp3T38YEU .marker.cross{stroke:#333333;}#mermaid-svg-cEUNC1ekp3T38YEU svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-cEUNC1ekp3T38YEU p{margin:0;}#mermaid-svg-cEUNC1ekp3T38YEU .edge{stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .section&#8211;1 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section&#8211;1 path,#mermaid-svg-cEUNC1ekp3T38YEU .section&#8211;1 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section&#8211;1 path{fill:hsl(240, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section&#8211;1 text{fill:#ffffff;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon&#8211;1{font-size:40px;color:#ffffff;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge&#8211;1{stroke:hsl(240, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth&#8211;1{stroke-width:17;}#mermaid-svg-cEUNC1ekp3T38YEU .section&#8211;1 line{stroke:hsl(60, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:#ffffff;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-0 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-0 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-0 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-0 path{fill:hsl(60, 100%, 73.5294117647%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-0 text{fill:black;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-0{font-size:40px;color:black;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-0{stroke:hsl(60, 100%, 73.5294117647%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-0{stroke-width:14;}#mermaid-svg-cEUNC1ekp3T38YEU .section-0 line{stroke:hsl(240, 100%, 83.5294117647%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:black;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-1 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-1 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-1 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-1 path{fill:hsl(80, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-1 text{fill:black;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-1{font-size:40px;color:black;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-1{stroke:hsl(80, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-1{stroke-width:11;}#mermaid-svg-cEUNC1ekp3T38YEU .section-1 line{stroke:hsl(260, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:black;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-2 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-2 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-2 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-2 path{fill:hsl(270, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-2 text{fill:#ffffff;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-2{font-size:40px;color:#ffffff;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-2{stroke:hsl(270, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-2{stroke-width:8;}#mermaid-svg-cEUNC1ekp3T38YEU .section-2 line{stroke:hsl(90, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:#ffffff;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-3 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-3 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-3 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-3 path{fill:hsl(300, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-3 text{fill:black;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-3{font-size:40px;color:black;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-3{stroke:hsl(300, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-3{stroke-width:5;}#mermaid-svg-cEUNC1ekp3T38YEU .section-3 line{stroke:hsl(120, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:black;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-4 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-4 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-4 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-4 path{fill:hsl(330, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-4 text{fill:black;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-4{font-size:40px;color:black;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-4{stroke:hsl(330, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-4{stroke-width:2;}#mermaid-svg-cEUNC1ekp3T38YEU .section-4 line{stroke:hsl(150, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:black;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-5 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-5 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-5 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-5 path{fill:hsl(0, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-5 text{fill:black;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-5{font-size:40px;color:black;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-5{stroke:hsl(0, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-5{stroke-width:-1;}#mermaid-svg-cEUNC1ekp3T38YEU .section-5 line{stroke:hsl(180, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:black;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-6 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-6 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-6 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-6 path{fill:hsl(30, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-6 text{fill:black;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-6{font-size:40px;color:black;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-6{stroke:hsl(30, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-6{stroke-width:-4;}#mermaid-svg-cEUNC1ekp3T38YEU .section-6 line{stroke:hsl(210, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:black;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-7 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-7 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-7 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-7 path{fill:hsl(90, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-7 text{fill:black;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-7{font-size:40px;color:black;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-7{stroke:hsl(90, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-7{stroke-width:-7;}#mermaid-svg-cEUNC1ekp3T38YEU .section-7 line{stroke:hsl(270, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:black;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-8 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-8 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-8 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-8 path{fill:hsl(150, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-8 text{fill:black;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-8{font-size:40px;color:black;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-8{stroke:hsl(150, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-8{stroke-width:-10;}#mermaid-svg-cEUNC1ekp3T38YEU .section-8 line{stroke:hsl(330, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:black;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-9 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-9 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-9 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-9 path{fill:hsl(180, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-9 text{fill:black;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-9{font-size:40px;color:black;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-9{stroke:hsl(180, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-9{stroke-width:-13;}#mermaid-svg-cEUNC1ekp3T38YEU .section-9 line{stroke:hsl(0, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:black;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-10 rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-10 path,#mermaid-svg-cEUNC1ekp3T38YEU .section-10 circle,#mermaid-svg-cEUNC1ekp3T38YEU .section-10 path{fill:hsl(210, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-10 text{fill:black;}#mermaid-svg-cEUNC1ekp3T38YEU .node-icon-10{font-size:40px;color:black;}#mermaid-svg-cEUNC1ekp3T38YEU .section-edge-10{stroke:hsl(210, 100%, 76.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .edge-depth-10{stroke-width:-16;}#mermaid-svg-cEUNC1ekp3T38YEU .section-10 line{stroke:hsl(30, 100%, 86.2745098039%);stroke-width:3;}#mermaid-svg-cEUNC1ekp3T38YEU .lineWrapper line{stroke:black;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled,#mermaid-svg-cEUNC1ekp3T38YEU .disabled circle,#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:lightgray;}#mermaid-svg-cEUNC1ekp3T38YEU .disabled text{fill:#efefef;}#mermaid-svg-cEUNC1ekp3T38YEU .section-root rect,#mermaid-svg-cEUNC1ekp3T38YEU .section-root path,#mermaid-svg-cEUNC1ekp3T38YEU .section-root circle{fill:hsl(240, 100%, 46.2745098039%);}#mermaid-svg-cEUNC1ekp3T38YEU .section-root text{fill:#ffffff;}#mermaid-svg-cEUNC1ekp3T38YEU .icon-container{height:100%;display:flex;justify-content:center;align-items:center;}#mermaid-svg-cEUNC1ekp3T38YEU .edge{fill:none;}#mermaid-svg-cEUNC1ekp3T38YEU .eventWrapper{filter:brightness(120%);}#mermaid-svg-cEUNC1ekp3T38YEU :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>        2022 <\/p>\n<p>        Naive RAG (\u57fa\u7840\u7248) <\/p>\n<p>        \u7b80\u5355\u68c0\u7d22&#043;\u751f\u6210 <\/p>\n<p>        2023 <\/p>\n<p>        Advanced RAG <\/p>\n<p>        (\u589e\u5f3a\u7248) <\/p>\n<p>        \u67e5\u8be2\u589e\u5f3a&#043;\u6df7\u5408\u68c0\u7d22&#043;\u91cd\u6392\u5e8f <\/p>\n<p>        2024 <\/p>\n<p>        Modular RAG <\/p>\n<p>        (\u6a21\u5757\u5316\u7248) <\/p>\n<p>        \u52a8\u6001\u8def\u7531&#043;\u81ea\u6211\u8bc4\u4f30 <\/p>\n<p>    RAG \u67b6\u6784\u6f14\u8fdb\u53f2 <\/p>\n<h4>1.2 \u4e09\u4ee3\u67b6\u6784\u8be6\u7ec6\u5bf9\u6bd4<\/h4>\n<table>\n<tr>\u7279\u6027\u7ef4\u5ea6Naive RAGAdvanced RAGModular RAG<\/tr>\n<tbody>\n<tr>\n<td align=\"left\">\u590d\u6742\u5ea6<\/td>\n<td align=\"left\">\u7b80\u5355<\/td>\n<td align=\"left\">\u4e2d\u7b49<\/td>\n<td align=\"left\">\u590d\u6742<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u68c0\u7d22\u7cbe\u5ea6<\/td>\n<td align=\"left\">\u57fa\u7840<\/td>\n<td align=\"left\">\u9ad8&#xff08;\u6df7\u5408&#043;\u91cd\u6392&#xff09;<\/td>\n<td align=\"left\">\u6781\u9ad8&#xff08;\u52a8\u6001\u4f18\u5316&#xff09;<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u7075\u6d3b\u6027<\/td>\n<td align=\"left\">\u56fa\u5b9a\u6d41\u7a0b<\/td>\n<td align=\"left\">\u53ef\u914d\u7f6e\u7b56\u7565<\/td>\n<td align=\"left\">\u5b8c\u5168\u6a21\u5757\u5316<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u9002\u7528\u573a\u666f<\/td>\n<td align=\"left\">\u7b80\u5355\u95ee\u7b54<\/td>\n<td align=\"left\">\u4f01\u4e1a\u7ea7\u5e94\u7528<\/td>\n<td align=\"left\">\u590d\u6742\u63a8\u7406\u4efb\u52a1<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h4>1.3 Naive RAG &#8211; \u57fa\u7840\u67b6\u6784<\/h4>\n<p>\u5de5\u4f5c\u539f\u7406&#xff1a;\u6700\u7b80\u5355\u7684RAG\u5b9e\u73b0&#xff0c;\u5c31\u50cf&#034;\u67e5\u5b57\u5178-\u7ffb\u8bd1&#034;\u7684\u8fc7\u7a0b&#xff1a;<\/p>\n<li>\u68c0\u7d22&#xff1a;\u5728\u5411\u91cf\u6570\u636e\u5e93\u4e2d\u67e5\u627e\u4e0e\u95ee\u9898\u76f8\u5173\u7684\u6587\u6863<\/li>\n<li>\u6784\u5efa&#xff1a;\u5c06\u68c0\u7d22\u5230\u7684\u6587\u6863\u548c\u95ee\u9898\u4e00\u8d77\u7ec4\u6210Prompt<\/li>\n<li>\u751f\u6210&#xff1a;\u8ba9LLM\u57fa\u4e8e\u8fd9\u4e9b\u4fe1\u606f\u751f\u6210\u7b54\u6848<\/li>\n<p>  #mermaid-svg-woA7l6vtMOin57Cp{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-woA7l6vtMOin57Cp .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-woA7l6vtMOin57Cp .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-woA7l6vtMOin57Cp .error-icon{fill:#552222;}#mermaid-svg-woA7l6vtMOin57Cp .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-woA7l6vtMOin57Cp .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-woA7l6vtMOin57Cp .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-woA7l6vtMOin57Cp .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-woA7l6vtMOin57Cp .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-woA7l6vtMOin57Cp .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-woA7l6vtMOin57Cp .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-woA7l6vtMOin57Cp .marker{fill:#333333;stroke:#333333;}#mermaid-svg-woA7l6vtMOin57Cp .marker.cross{stroke:#333333;}#mermaid-svg-woA7l6vtMOin57Cp svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-woA7l6vtMOin57Cp p{margin:0;}#mermaid-svg-woA7l6vtMOin57Cp .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-woA7l6vtMOin57Cp .cluster-label text{fill:#333;}#mermaid-svg-woA7l6vtMOin57Cp .cluster-label span{color:#333;}#mermaid-svg-woA7l6vtMOin57Cp .cluster-label span p{background-color:transparent;}#mermaid-svg-woA7l6vtMOin57Cp .label text,#mermaid-svg-woA7l6vtMOin57Cp span{fill:#333;color:#333;}#mermaid-svg-woA7l6vtMOin57Cp .node rect,#mermaid-svg-woA7l6vtMOin57Cp .node circle,#mermaid-svg-woA7l6vtMOin57Cp .node ellipse,#mermaid-svg-woA7l6vtMOin57Cp .node polygon,#mermaid-svg-woA7l6vtMOin57Cp .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-woA7l6vtMOin57Cp .rough-node .label text,#mermaid-svg-woA7l6vtMOin57Cp .node .label text,#mermaid-svg-woA7l6vtMOin57Cp .image-shape .label,#mermaid-svg-woA7l6vtMOin57Cp .icon-shape .label{text-anchor:middle;}#mermaid-svg-woA7l6vtMOin57Cp .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-woA7l6vtMOin57Cp .rough-node .label,#mermaid-svg-woA7l6vtMOin57Cp .node .label,#mermaid-svg-woA7l6vtMOin57Cp .image-shape .label,#mermaid-svg-woA7l6vtMOin57Cp .icon-shape .label{text-align:center;}#mermaid-svg-woA7l6vtMOin57Cp .node.clickable{cursor:pointer;}#mermaid-svg-woA7l6vtMOin57Cp .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-woA7l6vtMOin57Cp .arrowheadPath{fill:#333333;}#mermaid-svg-woA7l6vtMOin57Cp .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-woA7l6vtMOin57Cp .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-woA7l6vtMOin57Cp .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-woA7l6vtMOin57Cp .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-woA7l6vtMOin57Cp .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-woA7l6vtMOin57Cp .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-woA7l6vtMOin57Cp .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-woA7l6vtMOin57Cp .cluster text{fill:#333;}#mermaid-svg-woA7l6vtMOin57Cp .cluster span{color:#333;}#mermaid-svg-woA7l6vtMOin57Cp div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-woA7l6vtMOin57Cp .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-woA7l6vtMOin57Cp rect.text{fill:none;stroke-width:0;}#mermaid-svg-woA7l6vtMOin57Cp .icon-shape,#mermaid-svg-woA7l6vtMOin57Cp .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-woA7l6vtMOin57Cp .icon-shape p,#mermaid-svg-woA7l6vtMOin57Cp .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-woA7l6vtMOin57Cp .icon-shape rect,#mermaid-svg-woA7l6vtMOin57Cp .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-woA7l6vtMOin57Cp .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-woA7l6vtMOin57Cp .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-woA7l6vtMOin57Cp :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7528\u6237\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5411\u91cf\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u76f8\u5173\u6587\u6863 Top-K<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u6784\u5efaPrompt<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>LLM\u751f\u6210\u7b54\u6848<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u8fd4\u56de\u7ed3\u679c<\/p>\n<p><\/span> <\/p>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u5411\u91cf\u68c0\u7d22&#xff1a;\u5c06\u6587\u672c\u8f6c\u6362\u4e3a\u9ad8\u7ef4\u5411\u91cf&#xff0c;\u901a\u8fc7\u8ba1\u7b97\u5411\u91cf\u76f8\u4f3c\u5ea6\u627e\u5230\u76f8\u5173\u6587\u6863<\/li>\n<li>Top-K&#xff1a;\u68c0\u7d22\u6700\u76f8\u5173\u7684\u524dK\u4e2a\u6587\u6863&#xff0c;\u901a\u5e38K\u53d63-10<\/li>\n<li>Prompt Engineering&#xff1a;\u7cbe\u5fc3\u8bbe\u8ba1\u8f93\u5165\u7ed9LLM\u7684\u63d0\u793a\u8bcd&#xff0c;\u4ee5\u83b7\u5f97\u66f4\u597d\u7684\u8f93\u51fa<\/li>\n<\/ul>\n<h5>Python\u5b9e\u73b0\u793a\u4f8b<\/h5>\n<p><span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>embeddings <span class=\"token keyword\">import<\/span> OpenAIEmbeddings<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>vectorstores <span class=\"token keyword\">import<\/span> Chroma<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>chat_models <span class=\"token keyword\">import<\/span> ChatOpenAI<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>chains <span class=\"token keyword\">import<\/span> RetrievalQA<\/p>\n<p><span class=\"token comment\"># \u6b65\u9aa41: \u521d\u59cb\u5316\u5d4c\u5165\u6a21\u578b &#8211; \u5c06\u6587\u672c\u8f6c\u6362\u4e3a\u5411\u91cf<\/span><br \/>\nembeddings <span class=\"token operator\">&#061;<\/span> OpenAIEmbeddings<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u6b65\u9aa42: \u521b\u5efa\u5411\u91cf\u6570\u636e\u5e93 &#8211; \u5b58\u50a8\u548c\u68c0\u7d22\u6587\u6863\u5411\u91cf<\/span><br \/>\nvectorstore <span class=\"token operator\">&#061;<\/span> Chroma<span class=\"token punctuation\">(<\/span><br \/>\n    collection_name<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;knowledge_base&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    embedding_function<span class=\"token operator\">&#061;<\/span>embeddings<br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u6dfb\u52a0\u77e5\u8bc6\u6587\u6863<\/span><br \/>\ndocuments <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><br \/>\n    <span class=\"token string\">&#034;RAG\u662fRetrieval-Augmented Generation\u7684\u7f29\u5199&#8230;&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    <span class=\"token string\">&#034;\u5411\u91cf\u6570\u636e\u5e93\u662f\u4e13\u95e8\u5b58\u50a8\u548c\u68c0\u7d22\u5411\u91cf\u7684\u6570\u636e\u5e93&#8230;&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    <span class=\"token string\">&#034;Embedding\u6a21\u578b\u5c06\u6587\u672c\u8f6c\u6362\u4e3a\u9ad8\u7ef4\u5411\u91cf&#8230;&#034;<\/span><br \/>\n<span class=\"token punctuation\">]<\/span><br \/>\nvectorstore<span class=\"token punctuation\">.<\/span>add_documents<span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u6b65\u9aa43: \u521b\u5efaRAG\u94fe<\/span><br \/>\nretriever <span class=\"token operator\">&#061;<\/span> vectorstore<span class=\"token punctuation\">.<\/span>as_retriever<span class=\"token punctuation\">(<\/span>search_kwargs<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;k&#034;<\/span><span class=\"token punctuation\">:<\/span> <span class=\"token number\">3<\/span><span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><br \/>\nqa_chain <span class=\"token operator\">&#061;<\/span> RetrievalQA<span class=\"token punctuation\">.<\/span>from_chain_type<span class=\"token punctuation\">(<\/span><br \/>\n    llm<span class=\"token operator\">&#061;<\/span>ChatOpenAI<span class=\"token punctuation\">(<\/span>model<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;gpt-4&#034;<\/span><span class=\"token punctuation\">,<\/span> temperature<span class=\"token operator\">&#061;<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    retriever<span class=\"token operator\">&#061;<\/span>retriever<span class=\"token punctuation\">,<\/span><br \/>\n    return_source_documents<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u6267\u884c\u67e5\u8be2<\/span><br \/>\nresult <span class=\"token operator\">&#061;<\/span> qa_chain<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;query&#034;<\/span><span class=\"token punctuation\">:<\/span> <span class=\"token string\">&#034;\u4ec0\u4e48\u662fRAG\u6280\u672f&#xff1f;&#034;<\/span><span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<h4>1.4 Advanced RAG &#8211; \u589e\u5f3a\u67b6\u6784<\/h4>\n<p>\u6838\u5fc3\u75db\u70b9&#xff1a;Naive RAG\u7684\u4e09\u5927\u95ee\u9898<\/p>\n<ul>\n<li>\u8bed\u4e49\u9e3f\u6c9f&#xff1a;\u7528\u6237\u95ee\u9898\u548c\u6587\u6863\u7528\u8bcd\u4e0d\u540c&#xff0c;\u68c0\u7d22\u5931\u8d25<\/li>\n<li>\u7cbe\u5ea6\u4e0d\u8db3&#xff1a;\u5411\u91cf\u76f8\u4f3c\u5ea6\u4e0d\u7b49\u4e8e\u95ee\u9898\u76f8\u5173\u6027<\/li>\n<li>\u5355\u4e00\u7ef4\u5ea6&#xff1a;\u53ea\u7528\u8bed\u4e49\u68c0\u7d22&#xff0c;\u9519\u8fc7\u5173\u952e\u8bcd\u5339\u914d<\/li>\n<\/ul>\n<p>Advanced RAG\u89e3\u51b3\u65b9\u6848&#xff1a;<\/p>\n<p>  #mermaid-svg-0xn5cTjlm7cR6ChE{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-0xn5cTjlm7cR6ChE .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-0xn5cTjlm7cR6ChE .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-0xn5cTjlm7cR6ChE .error-icon{fill:#552222;}#mermaid-svg-0xn5cTjlm7cR6ChE .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-0xn5cTjlm7cR6ChE .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-0xn5cTjlm7cR6ChE .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-0xn5cTjlm7cR6ChE .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-0xn5cTjlm7cR6ChE .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-0xn5cTjlm7cR6ChE .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-0xn5cTjlm7cR6ChE .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-0xn5cTjlm7cR6ChE .marker{fill:#333333;stroke:#333333;}#mermaid-svg-0xn5cTjlm7cR6ChE .marker.cross{stroke:#333333;}#mermaid-svg-0xn5cTjlm7cR6ChE svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-0xn5cTjlm7cR6ChE p{margin:0;}#mermaid-svg-0xn5cTjlm7cR6ChE .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-0xn5cTjlm7cR6ChE .cluster-label text{fill:#333;}#mermaid-svg-0xn5cTjlm7cR6ChE .cluster-label span{color:#333;}#mermaid-svg-0xn5cTjlm7cR6ChE .cluster-label span p{background-color:transparent;}#mermaid-svg-0xn5cTjlm7cR6ChE .label text,#mermaid-svg-0xn5cTjlm7cR6ChE span{fill:#333;color:#333;}#mermaid-svg-0xn5cTjlm7cR6ChE .node rect,#mermaid-svg-0xn5cTjlm7cR6ChE .node circle,#mermaid-svg-0xn5cTjlm7cR6ChE .node ellipse,#mermaid-svg-0xn5cTjlm7cR6ChE .node polygon,#mermaid-svg-0xn5cTjlm7cR6ChE .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-0xn5cTjlm7cR6ChE .rough-node .label text,#mermaid-svg-0xn5cTjlm7cR6ChE .node .label text,#mermaid-svg-0xn5cTjlm7cR6ChE .image-shape .label,#mermaid-svg-0xn5cTjlm7cR6ChE .icon-shape .label{text-anchor:middle;}#mermaid-svg-0xn5cTjlm7cR6ChE .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-0xn5cTjlm7cR6ChE .rough-node .label,#mermaid-svg-0xn5cTjlm7cR6ChE .node .label,#mermaid-svg-0xn5cTjlm7cR6ChE .image-shape .label,#mermaid-svg-0xn5cTjlm7cR6ChE .icon-shape .label{text-align:center;}#mermaid-svg-0xn5cTjlm7cR6ChE .node.clickable{cursor:pointer;}#mermaid-svg-0xn5cTjlm7cR6ChE .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-0xn5cTjlm7cR6ChE .arrowheadPath{fill:#333333;}#mermaid-svg-0xn5cTjlm7cR6ChE .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-0xn5cTjlm7cR6ChE .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-0xn5cTjlm7cR6ChE .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-0xn5cTjlm7cR6ChE .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-0xn5cTjlm7cR6ChE .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-0xn5cTjlm7cR6ChE .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-0xn5cTjlm7cR6ChE .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-0xn5cTjlm7cR6ChE .cluster text{fill:#333;}#mermaid-svg-0xn5cTjlm7cR6ChE .cluster span{color:#333;}#mermaid-svg-0xn5cTjlm7cR6ChE div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-0xn5cTjlm7cR6ChE .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-0xn5cTjlm7cR6ChE rect.text{fill:none;stroke-width:0;}#mermaid-svg-0xn5cTjlm7cR6ChE .icon-shape,#mermaid-svg-0xn5cTjlm7cR6ChE .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-0xn5cTjlm7cR6ChE .icon-shape p,#mermaid-svg-0xn5cTjlm7cR6ChE .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-0xn5cTjlm7cR6ChE .icon-shape rect,#mermaid-svg-0xn5cTjlm7cR6ChE .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-0xn5cTjlm7cR6ChE .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-0xn5cTjlm7cR6ChE .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-0xn5cTjlm7cR6ChE :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7528\u6237\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u67e5\u8be2\u589e\u5f3a<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u6df7\u5408\u68c0\u7d22\u7a00\u758f&#043;\u5bc6\u96c6<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u91cd\u6392\u5e8f Re-rank<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u6784\u5efaPrompt<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>LLM\u751f\u6210<\/p>\n<p><\/span> <\/p>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u67e5\u8be2\u589e\u5f3a&#xff1a;\u4f7f\u7528HyDE\u7b49\u6280\u672f\u6539\u5199\u7528\u6237\u95ee\u9898&#xff0c;\u4f7f\u5176\u66f4\u6613\u4e8e\u68c0\u7d22<\/li>\n<li>\u6df7\u5408\u68c0\u7d22&#xff1a;\u7ed3\u5408\u7a00\u758f\u68c0\u7d22&#xff08;BM25&#xff09;\u548c\u5bc6\u96c6\u68c0\u7d22&#xff08;\u5411\u91cf&#xff09;&#xff0c;\u517c\u987e\u5173\u952e\u8bcd\u548c\u8bed\u4e49<\/li>\n<li>\u91cd\u6392\u5e8f&#xff1a;\u7528\u4e13\u95e8\u6a21\u578b\u5bf9\u68c0\u7d22\u7ed3\u679c\u8fdb\u884c\u7cbe\u6392&#xff0c;\u63d0\u5347Top\u7ed3\u679c\u8d28\u91cf<\/li>\n<\/ul>\n<h4>1.5 Modular RAG &#8211; \u6a21\u5757\u5316\u67b6\u6784<\/h4>\n<p>\u8bbe\u8ba1\u54f2\u5b66&#xff1a;\u5c06RAG\u62c6\u89e3\u4e3a\u53ef\u63d2\u62d4\u7684\u6a21\u5757&#xff0c;\u50cf\u642d\u79ef\u6728\u4e00\u6837\u81ea\u7531\u7ec4\u5408<\/p>\n<p>  #mermaid-svg-v0udIuTaje09TKuj{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-v0udIuTaje09TKuj .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-v0udIuTaje09TKuj .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-v0udIuTaje09TKuj .error-icon{fill:#552222;}#mermaid-svg-v0udIuTaje09TKuj .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-v0udIuTaje09TKuj .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-v0udIuTaje09TKuj .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-v0udIuTaje09TKuj .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-v0udIuTaje09TKuj .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-v0udIuTaje09TKuj .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-v0udIuTaje09TKuj .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-v0udIuTaje09TKuj .marker{fill:#333333;stroke:#333333;}#mermaid-svg-v0udIuTaje09TKuj .marker.cross{stroke:#333333;}#mermaid-svg-v0udIuTaje09TKuj svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-v0udIuTaje09TKuj p{margin:0;}#mermaid-svg-v0udIuTaje09TKuj .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-v0udIuTaje09TKuj .cluster-label text{fill:#333;}#mermaid-svg-v0udIuTaje09TKuj .cluster-label span{color:#333;}#mermaid-svg-v0udIuTaje09TKuj .cluster-label span p{background-color:transparent;}#mermaid-svg-v0udIuTaje09TKuj .label text,#mermaid-svg-v0udIuTaje09TKuj span{fill:#333;color:#333;}#mermaid-svg-v0udIuTaje09TKuj .node rect,#mermaid-svg-v0udIuTaje09TKuj .node circle,#mermaid-svg-v0udIuTaje09TKuj .node ellipse,#mermaid-svg-v0udIuTaje09TKuj .node polygon,#mermaid-svg-v0udIuTaje09TKuj .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-v0udIuTaje09TKuj .rough-node .label text,#mermaid-svg-v0udIuTaje09TKuj .node .label text,#mermaid-svg-v0udIuTaje09TKuj .image-shape .label,#mermaid-svg-v0udIuTaje09TKuj .icon-shape .label{text-anchor:middle;}#mermaid-svg-v0udIuTaje09TKuj .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-v0udIuTaje09TKuj .rough-node .label,#mermaid-svg-v0udIuTaje09TKuj .node .label,#mermaid-svg-v0udIuTaje09TKuj .image-shape .label,#mermaid-svg-v0udIuTaje09TKuj .icon-shape .label{text-align:center;}#mermaid-svg-v0udIuTaje09TKuj .node.clickable{cursor:pointer;}#mermaid-svg-v0udIuTaje09TKuj .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-v0udIuTaje09TKuj .arrowheadPath{fill:#333333;}#mermaid-svg-v0udIuTaje09TKuj .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-v0udIuTaje09TKuj .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-v0udIuTaje09TKuj .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-v0udIuTaje09TKuj .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-v0udIuTaje09TKuj .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-v0udIuTaje09TKuj .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-v0udIuTaje09TKuj .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-v0udIuTaje09TKuj .cluster text{fill:#333;}#mermaid-svg-v0udIuTaje09TKuj .cluster span{color:#333;}#mermaid-svg-v0udIuTaje09TKuj div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-v0udIuTaje09TKuj .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-v0udIuTaje09TKuj rect.text{fill:none;stroke-width:0;}#mermaid-svg-v0udIuTaje09TKuj .icon-shape,#mermaid-svg-v0udIuTaje09TKuj .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-v0udIuTaje09TKuj .icon-shape p,#mermaid-svg-v0udIuTaje09TKuj .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-v0udIuTaje09TKuj .icon-shape rect,#mermaid-svg-v0udIuTaje09TKuj .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-v0udIuTaje09TKuj .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-v0udIuTaje09TKuj .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-v0udIuTaje09TKuj :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/p>\n<p>\u7b80\u5355\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/p>\n<p>\u590d\u6742\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/p>\n<p>\u591a\u6b65\u63a8\u7406<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/p>\n<p>\u4e0d\u6ee1\u610f<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/p>\n<p>\u6ee1\u610f<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7528\u6237\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u667a\u80fd\u8def\u7531<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u76f4\u63a5\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u95ee\u9898\u62c6\u89e3<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u8fed\u4ee3\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u91cd\u6392\u5e8f<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u591a\u8def\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u8d28\u91cf\u8bc4\u4f30<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u8fd4\u56de\u7b54\u6848<\/p>\n<p><\/span> <\/p>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u667a\u80fd\u8def\u7531&#xff1a;\u6839\u636e\u95ee\u9898\u7c7b\u578b\u81ea\u52a8\u9009\u62e9\u6700\u4f73\u68c0\u7d22\u7b56\u7565<\/li>\n<li>\u95ee\u9898\u62c6\u89e3&#xff1a;\u5c06\u590d\u6742\u95ee\u9898\u5206\u89e3\u4e3a\u591a\u4e2a\u5b50\u95ee\u9898\u5206\u522b\u68c0\u7d22<\/li>\n<li>\u8fed\u4ee3\u68c0\u7d22&#xff1a;\u5229\u7528\u4e0a\u8f6e\u68c0\u7d22\u7ed3\u679c\u6539\u8fdb\u4e0b\u8f6e\u68c0\u7d22<\/li>\n<li>\u81ea\u6211\u8bc4\u4f30&#xff1a;\u7cfb\u7edf\u81ea\u52a8\u8bc4\u4f30\u68c0\u7d22\u8d28\u91cf\u5e76\u4f18\u5316<\/li>\n<\/ul>\n<hr \/>\n<h4>1.6 RAG \u5168\u94fe\u8def\u589e\u5f3a\u8def\u7ebf<\/h4>\n<p><img decoding=\"async\" src=\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260130114859-697c9aabec580.png\" alt=\"\u5728\u8fd9\u91cc\u63d2\u5165\u56fe\u7247\u63cf\u8ff0\" \/><\/p>\n<h3>\u4e8c\u3001\u68c0\u7d22\u524d\u589e\u5f3a<\/h3>\n<p>\u68c0\u7d22\u524d\u589e\u5f3a\u7684\u76ee\u6807\u662f\u6539\u8fdb\u7528\u6237\u7684\u67e5\u8be2&#xff0c;\u4f7f\u5176\u66f4\u5bb9\u6613\u88ab\u68c0\u7d22\u7cfb\u7edf\u7406\u89e3\u3002<\/p>\n<h4>2.1 HyDE (Hypothetical Document Embeddings)<\/h4>\n<p>  #mermaid-svg-tKdW5iiWpzpVzGkG{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-tKdW5iiWpzpVzGkG .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-tKdW5iiWpzpVzGkG .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-tKdW5iiWpzpVzGkG .error-icon{fill:#552222;}#mermaid-svg-tKdW5iiWpzpVzGkG .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-tKdW5iiWpzpVzGkG .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-tKdW5iiWpzpVzGkG .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-tKdW5iiWpzpVzGkG .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-tKdW5iiWpzpVzGkG .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-tKdW5iiWpzpVzGkG .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-tKdW5iiWpzpVzGkG .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-tKdW5iiWpzpVzGkG .marker{fill:#333333;stroke:#333333;}#mermaid-svg-tKdW5iiWpzpVzGkG .marker.cross{stroke:#333333;}#mermaid-svg-tKdW5iiWpzpVzGkG svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-tKdW5iiWpzpVzGkG p{margin:0;}#mermaid-svg-tKdW5iiWpzpVzGkG .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-tKdW5iiWpzpVzGkG .cluster-label text{fill:#333;}#mermaid-svg-tKdW5iiWpzpVzGkG .cluster-label span{color:#333;}#mermaid-svg-tKdW5iiWpzpVzGkG .cluster-label span p{background-color:transparent;}#mermaid-svg-tKdW5iiWpzpVzGkG .label text,#mermaid-svg-tKdW5iiWpzpVzGkG span{fill:#333;color:#333;}#mermaid-svg-tKdW5iiWpzpVzGkG .node rect,#mermaid-svg-tKdW5iiWpzpVzGkG .node circle,#mermaid-svg-tKdW5iiWpzpVzGkG .node ellipse,#mermaid-svg-tKdW5iiWpzpVzGkG .node polygon,#mermaid-svg-tKdW5iiWpzpVzGkG .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-tKdW5iiWpzpVzGkG .rough-node .label text,#mermaid-svg-tKdW5iiWpzpVzGkG .node .label text,#mermaid-svg-tKdW5iiWpzpVzGkG .image-shape .label,#mermaid-svg-tKdW5iiWpzpVzGkG .icon-shape .label{text-anchor:middle;}#mermaid-svg-tKdW5iiWpzpVzGkG .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-tKdW5iiWpzpVzGkG .rough-node .label,#mermaid-svg-tKdW5iiWpzpVzGkG .node .label,#mermaid-svg-tKdW5iiWpzpVzGkG .image-shape .label,#mermaid-svg-tKdW5iiWpzpVzGkG .icon-shape .label{text-align:center;}#mermaid-svg-tKdW5iiWpzpVzGkG .node.clickable{cursor:pointer;}#mermaid-svg-tKdW5iiWpzpVzGkG .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-tKdW5iiWpzpVzGkG .arrowheadPath{fill:#333333;}#mermaid-svg-tKdW5iiWpzpVzGkG .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-tKdW5iiWpzpVzGkG .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-tKdW5iiWpzpVzGkG .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-tKdW5iiWpzpVzGkG .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-tKdW5iiWpzpVzGkG .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-tKdW5iiWpzpVzGkG .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-tKdW5iiWpzpVzGkG .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-tKdW5iiWpzpVzGkG .cluster text{fill:#333;}#mermaid-svg-tKdW5iiWpzpVzGkG .cluster span{color:#333;}#mermaid-svg-tKdW5iiWpzpVzGkG div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-tKdW5iiWpzpVzGkG .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-tKdW5iiWpzpVzGkG rect.text{fill:none;stroke-width:0;}#mermaid-svg-tKdW5iiWpzpVzGkG .icon-shape,#mermaid-svg-tKdW5iiWpzpVzGkG .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-tKdW5iiWpzpVzGkG .icon-shape p,#mermaid-svg-tKdW5iiWpzpVzGkG .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-tKdW5iiWpzpVzGkG .icon-shape rect,#mermaid-svg-tKdW5iiWpzpVzGkG .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-tKdW5iiWpzpVzGkG .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-tKdW5iiWpzpVzGkG .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-tKdW5iiWpzpVzGkG :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7528\u6237\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>LLM\u751f\u6210\u5047\u8bbe\u7b54\u6848<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5047\u8bbe\u6027\u6587\u6863<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5411\u91cf\u7f16\u7801<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5411\u91cf\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u76f8\u5173\u6587\u6863<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u8fd4\u56de\u7ed3\u679c<\/p>\n<p><\/span> <\/p>\n<p>\u6838\u5fc3\u539f\u7406&#xff1a;\u7528\u6237\u7684\u95ee\u9898\u5f80\u5f80\u4e0d\u591f\u7cbe\u51c6\u3002HyDE\u901a\u8fc7\u751f\u6210\u5047\u8bbe\u6027\u7b54\u6848\u6765\u6539\u8fdb\u68c0\u7d22\u3002<\/p>\n<p>\u5de5\u4f5c\u539f\u7406<\/p>\n<ul>\n<li>\u539f\u95ee\u9898&#xff1a;\u201c\u624b\u673a\u7535\u6c60\u4e0d\u8010\u7528\u7684\u539f\u56e0\u201d<\/li>\n<li>\u751f\u6210\u5047\u8bbe\u7b54\u6848&#xff1a;\u201c\u624b\u673a\u7535\u6c60\u4e0d\u8010\u7528\u7684\u4e3b\u8981\u539f\u56e0\u5305\u62ec&#xff1a;\u540e\u53f0\u5e94\u7528\u8017\u7535\u3001\u5c4f\u5e55\u4eae\u5ea6\u8bbe\u7f6e\u8fc7\u9ad8\u3001\u7535\u6c60\u8001\u5316\u2026\u201d<\/li>\n<li>\u7528\u5047\u8bbe\u7b54\u6848\u68c0\u7d22&#xff1a;\u7cbe\u5ea6\u5927\u5e45\u63d0\u5347&#xff01;<\/li>\n<\/ul>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u5047\u8bbe\u6027\u6587\u6863&#xff1a;\u8ba9LLM\u751f\u6210\u4e00\u4e2a\u53ef\u80fd\u7684\u7b54\u6848&#xff0c;\u8fd9\u4e2a\u7b54\u6848\u5305\u542b\u95ee\u9898\u76f8\u5173\u7684\u5173\u952e\u8bcd\u548c\u8bed\u4e49<\/li>\n<li>\u5411\u91cf\u8868\u793a&#xff1a;\u5047\u8bbe\u7b54\u6848\u7684\u5411\u91cf\u8868\u793a\u4e0e\u771f\u5b9e\u7b54\u6848\u7684\u5411\u91cf\u8868\u793a\u66f4\u63a5\u8fd1<\/li>\n<\/ul>\n<p><span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>prompts <span class=\"token keyword\">import<\/span> PromptTemplate<\/p>\n<p><span class=\"token comment\"># HyDE Prompt<\/span><br \/>\nhyde_prompt <span class=\"token operator\">&#061;<\/span> PromptTemplate<span class=\"token punctuation\">(<\/span><br \/>\n    input_variables<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;question&#034;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    template<span class=\"token operator\">&#061;<\/span><span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u8bf7\u5199\u4e00\u4e2a\u53ef\u80fd\u7684\u7b54\u6848\u6765\u56de\u7b54\u8fd9\u4e2a\u95ee\u9898&#xff1a;{question}<\/p>\n<p>\u8981\u6c42&#xff1a;\u7b54\u6848\u5e94\u8be5\u50cf\u662f\u5728\u767e\u79d1\u5168\u4e66\u6216\u4e13\u4e1a\u6587\u6863\u4e2d\u627e\u5230\u7684\u90a3\u6837\u3002&#034;&#034;&#034;<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token keyword\">def<\/span> <span class=\"token function\">hyde_retrieval<\/span><span class=\"token punctuation\">(<\/span>question<span class=\"token punctuation\">,<\/span> vectorstore<span class=\"token punctuation\">,<\/span> top_k<span class=\"token operator\">&#061;<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token comment\"># \u751f\u6210\u5047\u8bbe\u6027\u6587\u6863<\/span><br \/>\n    hypothetical_doc <span class=\"token operator\">&#061;<\/span> llm<span class=\"token punctuation\">(<\/span>hyde_prompt<span class=\"token punctuation\">.<\/span><span class=\"token builtin\">format<\/span><span class=\"token punctuation\">(<\/span>question<span class=\"token operator\">&#061;<\/span>question<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u7528\u5047\u8bbe\u6587\u6863\u8fdb\u884c\u5411\u91cf\u68c0\u7d22<\/span><br \/>\n    retriever <span class=\"token operator\">&#061;<\/span> vectorstore<span class=\"token punctuation\">.<\/span>as_retriever<span class=\"token punctuation\">(<\/span>search_kwargs<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;k&#034;<\/span><span class=\"token punctuation\">:<\/span> top_k<span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">return<\/span> retriever<span class=\"token punctuation\">.<\/span>get_relevant_documents<span class=\"token punctuation\">(<\/span>hypothetical_doc<span class=\"token punctuation\">)<\/span><\/p>\n<h4>2.2 Query2Doc<\/h4>\n<p>  #mermaid-svg-kxjd84eVN5T4JhZ4{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-kxjd84eVN5T4JhZ4 .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-kxjd84eVN5T4JhZ4 .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-kxjd84eVN5T4JhZ4 .error-icon{fill:#552222;}#mermaid-svg-kxjd84eVN5T4JhZ4 .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-kxjd84eVN5T4JhZ4 .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-kxjd84eVN5T4JhZ4 .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-kxjd84eVN5T4JhZ4 .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-kxjd84eVN5T4JhZ4 .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-kxjd84eVN5T4JhZ4 .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-kxjd84eVN5T4JhZ4 .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-kxjd84eVN5T4JhZ4 .marker{fill:#333333;stroke:#333333;}#mermaid-svg-kxjd84eVN5T4JhZ4 .marker.cross{stroke:#333333;}#mermaid-svg-kxjd84eVN5T4JhZ4 svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-kxjd84eVN5T4JhZ4 p{margin:0;}#mermaid-svg-kxjd84eVN5T4JhZ4 .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-kxjd84eVN5T4JhZ4 .cluster-label text{fill:#333;}#mermaid-svg-kxjd84eVN5T4JhZ4 .cluster-label span{color:#333;}#mermaid-svg-kxjd84eVN5T4JhZ4 .cluster-label span p{background-color:transparent;}#mermaid-svg-kxjd84eVN5T4JhZ4 .label text,#mermaid-svg-kxjd84eVN5T4JhZ4 span{fill:#333;color:#333;}#mermaid-svg-kxjd84eVN5T4JhZ4 .node rect,#mermaid-svg-kxjd84eVN5T4JhZ4 .node circle,#mermaid-svg-kxjd84eVN5T4JhZ4 .node ellipse,#mermaid-svg-kxjd84eVN5T4JhZ4 .node polygon,#mermaid-svg-kxjd84eVN5T4JhZ4 .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-kxjd84eVN5T4JhZ4 .rough-node .label text,#mermaid-svg-kxjd84eVN5T4JhZ4 .node .label text,#mermaid-svg-kxjd84eVN5T4JhZ4 .image-shape .label,#mermaid-svg-kxjd84eVN5T4JhZ4 .icon-shape .label{text-anchor:middle;}#mermaid-svg-kxjd84eVN5T4JhZ4 .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-kxjd84eVN5T4JhZ4 .rough-node .label,#mermaid-svg-kxjd84eVN5T4JhZ4 .node .label,#mermaid-svg-kxjd84eVN5T4JhZ4 .image-shape .label,#mermaid-svg-kxjd84eVN5T4JhZ4 .icon-shape .label{text-align:center;}#mermaid-svg-kxjd84eVN5T4JhZ4 .node.clickable{cursor:pointer;}#mermaid-svg-kxjd84eVN5T4JhZ4 .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-kxjd84eVN5T4JhZ4 .arrowheadPath{fill:#333333;}#mermaid-svg-kxjd84eVN5T4JhZ4 .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-kxjd84eVN5T4JhZ4 .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-kxjd84eVN5T4JhZ4 .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-kxjd84eVN5T4JhZ4 .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-kxjd84eVN5T4JhZ4 .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-kxjd84eVN5T4JhZ4 .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-kxjd84eVN5T4JhZ4 .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-kxjd84eVN5T4JhZ4 .cluster text{fill:#333;}#mermaid-svg-kxjd84eVN5T4JhZ4 .cluster span{color:#333;}#mermaid-svg-kxjd84eVN5T4JhZ4 div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-kxjd84eVN5T4JhZ4 .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-kxjd84eVN5T4JhZ4 rect.text{fill:none;stroke-width:0;}#mermaid-svg-kxjd84eVN5T4JhZ4 .icon-shape,#mermaid-svg-kxjd84eVN5T4JhZ4 .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-kxjd84eVN5T4JhZ4 .icon-shape p,#mermaid-svg-kxjd84eVN5T4JhZ4 .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-kxjd84eVN5T4JhZ4 .icon-shape rect,#mermaid-svg-kxjd84eVN5T4JhZ4 .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-kxjd84eVN5T4JhZ4 .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-kxjd84eVN5T4JhZ4 .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-kxjd84eVN5T4JhZ4 :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7528\u6237\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>LLM\u751f\u6210\u6587\u6863\u7247\u6bb5<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u6587\u6863\u7247\u6bb550-100\u5b57<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5411\u91cf\u7f16\u7801<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5411\u91cf\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u76f8\u5173\u6587\u6863<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u8fd4\u56de\u7ed3\u679c<\/p>\n<p><\/span> <\/p>\n<p>\u6838\u5fc3\u539f\u7406&#xff1a;\u4e0eHyDE\u7c7b\u4f3c&#xff0c;\u4f46Query2Doc\u4e13\u6ce8\u4e8e\u751f\u6210\u66f4\u77ed\u7684\u6587\u6863\u7247\u6bb5\u6765\u6539\u8fdb\u68c0\u7d22\u3002<\/p>\n<p>Query2Doc vs HyDE<\/p>\n<table>\n<tr>\u7279\u6027HyDEQuery2Doc<\/tr>\n<tbody>\n<tr>\n<td>\u8f93\u51fa\u957f\u5ea6<\/td>\n<td>\u8f83\u957f&#xff08;\u5b8c\u6574\u7b54\u6848&#xff09;<\/td>\n<td>\u8f83\u77ed&#xff08;\u5173\u952e\u7247\u6bb5&#xff09;<\/td>\n<\/tr>\n<tr>\n<td>\u751f\u6210\u91cd\u70b9<\/td>\n<td>\u5b8c\u6574\u56de\u7b54\u95ee\u9898<\/td>\n<td>\u6355\u6349\u5173\u952e\u6982\u5ff5<\/td>\n<\/tr>\n<tr>\n<td>\u9002\u7528\u573a\u666f<\/td>\n<td>\u95ee\u7b54\u7cfb\u7edf<\/td>\n<td>\u6587\u6863\u68c0\u7d22<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u5de5\u4f5c\u539f\u7406&#xff1a;<\/p>\n<li>\u5c06\u7528\u6237\u95ee\u9898\u8f6c\u6362\u4e3a\u66f4\u4e30\u5bcc\u7684\u6587\u6863\u8868\u793a<\/li>\n<li>\u751f\u6210\u7684\u6587\u6863\u5305\u542b\u95ee\u9898\u76f8\u5173\u7684\u5173\u952e\u8bcd\u548c\u6982\u5ff5<\/li>\n<li>\u7528\u8fd9\u4e2a\u6587\u6863\u8fdb\u884c\u5411\u91cf\u68c0\u7d22<\/li>\n<p><span class=\"token comment\"># Query2Doc Prompt<\/span><br \/>\nquery2doc_prompt <span class=\"token operator\">&#061;<\/span> PromptTemplate<span class=\"token punctuation\">(<\/span><br \/>\n    input_variables<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;question&#034;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    template<span class=\"token operator\">&#061;<\/span><span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u57fa\u4e8e\u4ee5\u4e0b\u95ee\u9898&#xff0c;\u751f\u6210\u4e00\u4e2a\u7b80\u77ed\u7684\u6587\u6863\u7247\u6bb5&#xff08;50-100\u5b57&#xff09;\u3002<\/p>\n<p>\u8981\u6c42&#xff1a;<br \/>\n1. \u5305\u542b\u95ee\u9898\u4e2d\u7684\u5173\u952e\u6982\u5ff5<br \/>\n2. \u4f7f\u7528\u4e13\u4e1a\u672f\u8bed<br \/>\n3. \u4fdd\u6301\u7b80\u6d01<\/p>\n<p>\u95ee\u9898: {question}<\/p>\n<p>\u6587\u6863\u7247\u6bb5:&#034;&#034;&#034;<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token keyword\">def<\/span> <span class=\"token function\">query2doc_retrieval<\/span><span class=\"token punctuation\">(<\/span>question<span class=\"token punctuation\">,<\/span> vectorstore<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token comment\"># \u751f\u6210\u6587\u6863\u7247\u6bb5<\/span><br \/>\n    doc_fragment <span class=\"token operator\">&#061;<\/span> llm<span class=\"token punctuation\">(<\/span>query2doc_prompt<span class=\"token punctuation\">.<\/span><span class=\"token builtin\">format<\/span><span class=\"token punctuation\">(<\/span>question<span class=\"token operator\">&#061;<\/span>question<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u7528\u6587\u6863\u7247\u6bb5\u8fdb\u884c\u68c0\u7d22<\/span><br \/>\n    retriever <span class=\"token operator\">&#061;<\/span> vectorstore<span class=\"token punctuation\">.<\/span>as_retriever<span class=\"token punctuation\">(<\/span>search_kwargs<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;k&#034;<\/span><span class=\"token punctuation\">:<\/span> <span class=\"token number\">3<\/span><span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">return<\/span> retriever<span class=\"token punctuation\">.<\/span>get_relevant_documents<span class=\"token punctuation\">(<\/span>doc_fragment<span class=\"token punctuation\">)<\/span><\/p>\n<h4>2.3 \u67e5\u8be2\u91cd\u5199 (Query Rewriting)<\/h4>\n<p>\u6838\u5fc3\u539f\u7406&#xff1a;\u5c06\u53e3\u8bed\u5316\u3001\u6a21\u7cca\u7684\u7528\u6237\u95ee\u9898\u6539\u5199\u4e3a\u4e13\u4e1a\u3001\u7cbe\u786e\u7684\u5f62\u5f0f\u3002<\/p>\n<p><span class=\"token comment\"># \u67e5\u8be2\u91cd\u5199Prompt<\/span><br \/>\nrewrite_prompt <span class=\"token operator\">&#061;<\/span> PromptTemplate<span class=\"token punctuation\">(<\/span><br \/>\n    template<span class=\"token operator\">&#061;<\/span><span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u5c06\u4ee5\u4e0b\u7528\u6237\u95ee\u9898\u6539\u5199\u4e3a\u66f4\u9002\u5408\u641c\u7d22\u7684\u5f62\u5f0f\u3002<br \/>\n\u8981\u6c42&#xff1a;\u4fdd\u7559\u6838\u5fc3\u610f\u56fe&#xff0c;\u4f7f\u7528\u66f4\u4e13\u4e1a\u7684\u672f\u8bed&#xff0c;\u8865\u5145\u53ef\u80fd\u7684\u540c\u4e49\u8bcd\u3002<\/p>\n<p>\u539f\u95ee\u9898: {question}<\/p>\n<p>\u6539\u5199\u540e\u7684\u95ee\u9898:&#034;&#034;&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    input_variables<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;question&#034;<\/span><span class=\"token punctuation\">]<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u793a\u4f8b<\/span><br \/>\n<span class=\"token comment\"># \u539f\u95ee\u9898: &#034;\u624b\u673a\u8001\u6ca1\u7535\u548b\u529e&#034;<\/span><br \/>\n<span class=\"token comment\"># \u6539\u5199\u540e: &#034;\u624b\u673a\u7535\u6c60\u7eed\u822a\u65f6\u95f4\u77ed&#xff0c;\u5982\u4f55\u89e3\u51b3\u7535\u6c60\u6d88\u8017\u8fc7\u5feb\u7684\u95ee\u9898&#034;<\/span><\/p>\n<h4>2.4 \u67e5\u8be2\u5206\u89e3 (Query Decomposition)<\/h4>\n<p>\u6838\u5fc3\u539f\u7406&#xff1a;\u5c06\u590d\u6742\u7684\u590d\u5408\u95ee\u9898\u5206\u89e3\u4e3a\u591a\u4e2a\u7b80\u5355\u7684\u5b50\u95ee\u9898&#xff0c;\u5206\u522b\u68c0\u7d22\u540e\u518d\u5408\u5e76\u7ed3\u679c\u3002<\/p>\n<p>\u4e3a\u4ec0\u4e48\u9700\u8981\u67e5\u8be2\u5206\u89e3&#xff1f;<\/p>\n<p>\u7528\u6237\u95ee\u9898&#xff1a;\u201cRAG\u6280\u672f\u6709\u54ea\u4e9b\u4f18\u52bf&#xff1f;\u5728\u5b9e\u9645\u9879\u76ee\u4e2d\u5982\u4f55\u5e94\u7528&#xff1f;\u6709\u54ea\u4e9b\u6311\u6218&#xff1f;\u201d<\/p>\n<p>\u8fd9\u4e2a\u95ee\u9898\u5305\u542b3\u4e2a\u5b50\u95ee\u9898&#xff0c;\u5982\u679c\u76f4\u63a5\u68c0\u7d22\u53ef\u80fd\u627e\u4e0d\u5230\u540c\u65f6\u6db5\u76d6\u6240\u6709\u5185\u5bb9\u7684\u6587\u6863\u3002<\/p>\n<p>\u89e3\u51b3\u65b9\u6848&#xff1a;<\/p>\n<li>\u5206\u89e3\u4e3a3\u4e2a\u5b50\u95ee\u9898<\/li>\n<li>\u5206\u522b\u68c0\u7d22\u6bcf\u4e2a\u5b50\u95ee\u9898\u7684\u76f8\u5173\u6587\u6863<\/li>\n<li>\u5408\u5e76\u6240\u6709\u68c0\u7d22\u7ed3\u679c<\/li>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u590d\u5408\u95ee\u9898&#xff1a;\u5305\u542b\u591a\u4e2a\u5b50\u95ee\u9898\u6216\u591a\u4e2a\u65b9\u9762\u7684\u95ee\u9898<\/li>\n<li>\u5b50\u95ee\u9898\u72ec\u7acb\u68c0\u7d22&#xff1a;\u6bcf\u4e2a\u5b50\u95ee\u9898\u5355\u72ec\u8fdb\u884c\u68c0\u7d22&#xff0c;\u63d0\u5347\u53ec\u56de\u7387<\/li>\n<\/ul>\n<p><span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>prompts <span class=\"token keyword\">import<\/span> PromptTemplate<\/p>\n<p><span class=\"token comment\"># \u67e5\u8be2\u5206\u89e3Prompt<\/span><br \/>\ndecompose_prompt <span class=\"token operator\">&#061;<\/span> PromptTemplate<span class=\"token punctuation\">(<\/span><br \/>\n    template<span class=\"token operator\">&#061;<\/span><span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u5c06\u4ee5\u4e0b\u590d\u6742\u95ee\u9898\u5206\u89e3\u4e3a\u591a\u4e2a\u72ec\u7acb\u7684\u5b50\u95ee\u9898\u3002<\/p>\n<p>\u8981\u6c42:<br \/>\n1. \u6bcf\u4e2a\u5b50\u95ee\u9898\u90fd\u5e94\u8be5\u80fd\u72ec\u7acb\u68c0\u7d22<br \/>\n2. \u5b50\u95ee\u9898\u4e4b\u95f4\u5c3d\u91cf\u4e0d\u91cd\u590d<br \/>\n3. \u5206\u89e33-5\u4e2a\u5b50\u95ee\u9898<br \/>\n4. \u6bcf\u4e2a\u5b50\u95ee\u9898\u72ec\u7acb\u6210\u884c<\/p>\n<p>\u539f\u59cb\u95ee\u9898: {question}<\/p>\n<p>\u5b50\u95ee\u9898\u5217\u8868&#xff08;\u6bcf\u884c\u4e00\u4e2a&#xff09;:&#034;&#034;&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    input_variables<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;question&#034;<\/span><span class=\"token punctuation\">]<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token keyword\">def<\/span> <span class=\"token function\">decompose_query<\/span><span class=\"token punctuation\">(<\/span>question<span class=\"token punctuation\">,<\/span> vectorstore<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u67e5\u8be2\u5206\u89e3\u5e76\u68c0\u7d22&#034;&#034;&#034;<\/span><br \/>\n    <span class=\"token comment\"># 1. \u5206\u89e3\u95ee\u9898<\/span><br \/>\n    sub_questions_str <span class=\"token operator\">&#061;<\/span> llm<span class=\"token punctuation\">(<\/span>decompose_prompt<span class=\"token punctuation\">.<\/span><span class=\"token builtin\">format<\/span><span class=\"token punctuation\">(<\/span>question<span class=\"token operator\">&#061;<\/span>question<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    sub_questions <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span>q<span class=\"token punctuation\">.<\/span>strip<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token keyword\">for<\/span> q <span class=\"token keyword\">in<\/span> sub_questions_str<span class=\"token punctuation\">.<\/span>split<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;\\\\n&#039;<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token keyword\">if<\/span> q<span class=\"token punctuation\">.<\/span>strip<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p>    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5206\u89e3\u51fa <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span><span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span>sub_questions<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\"> \u4e2a\u5b50\u95ee\u9898:&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">for<\/span> i<span class=\"token punctuation\">,<\/span> sq <span class=\"token keyword\">in<\/span> <span class=\"token builtin\">enumerate<\/span><span class=\"token punctuation\">(<\/span>sub_questions<span class=\"token punctuation\">,<\/span> <span class=\"token number\">1<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;  <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>i<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">. <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>sq<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># 2. \u5bf9\u6bcf\u4e2a\u5b50\u95ee\u9898\u68c0\u7d22<\/span><br \/>\n    all_docs <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">]<\/span><br \/>\n    retriever <span class=\"token operator\">&#061;<\/span> vectorstore<span class=\"token punctuation\">.<\/span>as_retriever<span class=\"token punctuation\">(<\/span>search_kwargs<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;k&#034;<\/span><span class=\"token punctuation\">:<\/span> <span class=\"token number\">2<\/span><span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token keyword\">for<\/span> sq <span class=\"token keyword\">in<\/span> sub_questions<span class=\"token punctuation\">:<\/span><br \/>\n        docs <span class=\"token operator\">&#061;<\/span> retriever<span class=\"token punctuation\">.<\/span>get_relevant_documents<span class=\"token punctuation\">(<\/span>sq<span class=\"token punctuation\">)<\/span><br \/>\n        all_docs<span class=\"token punctuation\">.<\/span>extend<span class=\"token punctuation\">(<\/span>docs<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># 3. \u53bb\u91cd\u5e76\u8fd4\u56deTop-N<\/span><br \/>\n    unique_docs <span class=\"token operator\">&#061;<\/span> <span class=\"token builtin\">list<\/span><span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">{<\/span>doc<span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">:<\/span> doc <span class=\"token keyword\">for<\/span> doc <span class=\"token keyword\">in<\/span> all_docs<span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">.<\/span>values<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">return<\/span> unique_docs<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token number\">5<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p><span class=\"token comment\"># \u4f7f\u7528\u793a\u4f8b<\/span><br \/>\ncomplex_question <span class=\"token operator\">&#061;<\/span> <span class=\"token string\">&#034;RAG\u6280\u672f\u6709\u54ea\u4e9b\u4f18\u52bf&#xff1f;\u5728\u5b9e\u9645\u9879\u76ee\u4e2d\u5982\u4f55\u5e94\u7528&#xff1f;&#034;<\/span><br \/>\ndocs <span class=\"token operator\">&#061;<\/span> decompose_query<span class=\"token punctuation\">(<\/span>complex_question<span class=\"token punctuation\">,<\/span> vectorstore<span class=\"token punctuation\">)<\/span><\/p>\n<h4>2.5 Take Step Back<\/h4>\n<p>\u6838\u5fc3\u539f\u7406&#xff1a;\u8ba9AI&#034;\u9000\u4e00\u6b65&#034;\u601d\u8003&#xff0c;\u4ece\u66f4\u9ad8\u5c42\u6b21\u7684\u6982\u5ff5\u6765\u7406\u89e3\u95ee\u9898&#xff0c;\u7136\u540e\u518d\u8fdb\u884c\u68c0\u7d22\u3002<\/p>\n<p>Take Step Back \u793a\u4f8b<\/p>\n<p>\u7528\u6237\u95ee\u9898&#xff1a;\u201cPython\u4e2d\u7684\u88c5\u9970\u5668\u662f\u600e\u4e48\u5de5\u4f5c\u7684&#xff1f;\u5728\u5f02\u6b65\u7f16\u7a0b\u4e2d\u5982\u4f55\u4f7f\u7528&#xff1f;\u201d<\/p>\n<p>Step Back\u540e&#xff1a;\u201cPython\u7f16\u7a0b\u4e2d\u7684\u88c5\u9970\u5668\u6a21\u5f0f\u548c\u5f02\u6b65\u7f16\u7a0b\u6982\u5ff5\u201d<\/p>\n<p>\u4f18\u52bf&#xff1a;\u4ece\u66f4\u9ad8\u5c42\u6b21\u7406\u89e3\u95ee\u9898\u7684\u6838\u5fc3\u6982\u5ff5&#xff0c;\u68c0\u7d22\u66f4\u5168\u9762<\/p>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u62bd\u8c61\u5c42\u6b21&#xff1a;\u4ece\u5177\u4f53\u95ee\u9898\u4e0a\u5347\u5230\u66f4\u9ad8\u5c42\u6b21\u7684\u6982\u5ff5<\/li>\n<li>\u6982\u5ff5\u5173\u8054&#xff1a;\u901a\u8fc7\u9ad8\u5c42\u6b21\u6982\u5ff5\u5173\u8054\u66f4\u591a\u76f8\u5173\u6587\u6863<\/li>\n<\/ul>\n<p><span class=\"token comment\"># Take Step Back Prompt<\/span><br \/>\nstep_back_prompt <span class=\"token operator\">&#061;<\/span> PromptTemplate<span class=\"token punctuation\">(<\/span><br \/>\n    template<span class=\"token operator\">&#061;<\/span><span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u4f60\u662f\u4e00\u4e2a\u64c5\u957f\u6982\u5ff5\u62bd\u8c61\u7684AI\u52a9\u624b\u3002<\/p>\n<p>\u8bf7\u5c06\u4ee5\u4e0b\u5177\u4f53\u95ee\u9898\u62bd\u8c61\u4e3a\u4e00\u4e2a\u66f4\u9ad8\u5c42\u6b21\u7684\u6982\u5ff5\u6027\u95ee\u9898\u3002<\/p>\n<p>\u8981\u6c42&#xff1a;<br \/>\n1. \u63d0\u53d6\u95ee\u9898\u4e2d\u7684\u6838\u5fc3\u6982\u5ff5<br \/>\n2. \u4ece\u66f4\u9ad8\u5c42\u6b21\u91cd\u65b0\u8868\u8ff0\u95ee\u9898<br \/>\n3. \u4fdd\u6301\u95ee\u9898\u7684\u8bed\u4e49\u8303\u56f4<\/p>\n<p>\u539f\u59cb\u95ee\u9898: {question}<\/p>\n<p>\u62bd\u8c61\u540e\u7684\u95ee\u9898:&#034;&#034;&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    input_variables<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;question&#034;<\/span><span class=\"token punctuation\">]<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token keyword\">def<\/span> <span class=\"token function\">take_step_back_retrieval<\/span><span class=\"token punctuation\">(<\/span>question<span class=\"token punctuation\">,<\/span> vectorstore<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token triple-quoted-string string\">&#034;&#034;&#034;Step Back\u68c0\u7d22&#034;&#034;&#034;<\/span><br \/>\n    <span class=\"token comment\"># 1. \u751f\u6210\u62bd\u8c61\u95ee\u9898<\/span><br \/>\n    abstract_question <span class=\"token operator\">&#061;<\/span> llm<span class=\"token punctuation\">(<\/span>step_back_prompt<span class=\"token punctuation\">.<\/span><span class=\"token builtin\">format<\/span><span class=\"token punctuation\">(<\/span>question<span class=\"token operator\">&#061;<\/span>question<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u539f\u59cb\u95ee\u9898: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>question<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u62bd\u8c61\u95ee\u9898: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>abstract_question<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># 2. \u7528\u62bd\u8c61\u95ee\u9898\u68c0\u7d22<\/span><br \/>\n    retriever <span class=\"token operator\">&#061;<\/span> vectorstore<span class=\"token punctuation\">.<\/span>as_retriever<span class=\"token punctuation\">(<\/span>search_kwargs<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;k&#034;<\/span><span class=\"token punctuation\">:<\/span> <span class=\"token number\">3<\/span><span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    abstract_docs <span class=\"token operator\">&#061;<\/span> retriever<span class=\"token punctuation\">.<\/span>get_relevant_documents<span class=\"token punctuation\">(<\/span>abstract_question<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># 3. \u7528\u539f\u59cb\u95ee\u9898\u68c0\u7d22<\/span><br \/>\n    original_docs <span class=\"token operator\">&#061;<\/span> retriever<span class=\"token punctuation\">.<\/span>get_relevant_documents<span class=\"token punctuation\">(<\/span>question<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># 4. \u5408\u5e76\u53bb\u91cd<\/span><br \/>\n    all_docs <span class=\"token operator\">&#061;<\/span> abstract_docs <span class=\"token operator\">&#043;<\/span> original_docs<br \/>\n    unique_docs <span class=\"token operator\">&#061;<\/span> <span class=\"token builtin\">list<\/span><span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">{<\/span>doc<span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">:<\/span> doc <span class=\"token keyword\">for<\/span> doc <span class=\"token keyword\">in<\/span> all_docs<span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">.<\/span>values<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token keyword\">return<\/span> unique_docs<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token number\">5<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p><span class=\"token comment\"># \u4f7f\u7528\u793a\u4f8b<\/span><br \/>\nquestion <span class=\"token operator\">&#061;<\/span> <span class=\"token string\">&#034;Python\u4e2d\u7684\u88c5\u9970\u5668\u662f\u600e\u4e48\u5de5\u4f5c\u7684&#xff1f;\u5728\u5f02\u6b65\u7f16\u7a0b\u4e2d\u5982\u4f55\u4f7f\u7528&#xff1f;&#034;<\/span><br \/>\ndocs <span class=\"token operator\">&#061;<\/span> take_step_back_retrieval<span class=\"token punctuation\">(<\/span>question<span class=\"token punctuation\">,<\/span> vectorstore<span class=\"token punctuation\">)<\/span><\/p>\n<p>Take Step Back vs \u67e5\u8be2\u5206\u89e3&#xff1a;<\/p>\n<table>\n<tr>\u7279\u6027Take Step Back\u67e5\u8be2\u5206\u89e3<\/tr>\n<tbody>\n<tr>\n<td>\u5904\u7406\u65b9\u5f0f<\/td>\n<td>\u62bd\u8c61\u5230\u66f4\u9ad8\u5c42\u6b21<\/td>\n<td>\u5206\u89e3\u4e3a\u591a\u4e2a\u5b50\u95ee\u9898<\/td>\n<\/tr>\n<tr>\n<td>\u68c0\u7d22\u7b56\u7565<\/td>\n<td>\u7528\u62bd\u8c61\u95ee\u9898&#043;\u539f\u95ee\u9898<\/td>\n<td>\u5206\u522b\u68c0\u7d22\u6bcf\u4e2a\u5b50\u95ee\u9898<\/td>\n<\/tr>\n<tr>\n<td>\u9002\u7528\u573a\u666f<\/td>\n<td>\u9700\u8981\u5168\u5c40\u89c6\u89d2\u7684\u95ee\u9898<\/td>\n<td>\u591a\u65b9\u9762\u7684\u590d\u5408\u95ee\u9898<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<hr \/>\n<h3>\u4e09\u3001\u68c0\u7d22\u589e\u5f3a<\/h3>\n<p>\u68c0\u7d22\u589e\u5f3a\u7684\u76ee\u6807\u662f\u4f7f\u7528\u591a\u79cd\u7d22\u5f15\u548c\u68c0\u7d22\u7b56\u7565&#xff0c;\u63d0\u5347\u68c0\u7d22\u7684\u53ec\u56de\u7387\u548c\u51c6\u786e\u7387\u3002<\/p>\n<h4>3.1 \u6838\u5fc3\u6982\u5ff5<\/h4>\n<p>\u591a\u7d22\u5f15\u68c0\u7d22\u7b56\u7565\u7684\u6838\u5fc3\u601d\u60f3\u662f&#xff1a;\u4e0d\u540c\u7684\u7d22\u5f15\u7ef4\u5ea6\u6355\u83b7\u4e0d\u540c\u7684\u4fe1\u606f\u7279\u5f81\u3002<\/p>\n<p>  #mermaid-svg-xVSYOQeEx7E2tROM{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-xVSYOQeEx7E2tROM .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-xVSYOQeEx7E2tROM .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-xVSYOQeEx7E2tROM .error-icon{fill:#552222;}#mermaid-svg-xVSYOQeEx7E2tROM .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-xVSYOQeEx7E2tROM .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-xVSYOQeEx7E2tROM .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-xVSYOQeEx7E2tROM .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-xVSYOQeEx7E2tROM .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-xVSYOQeEx7E2tROM .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-xVSYOQeEx7E2tROM .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-xVSYOQeEx7E2tROM .marker{fill:#333333;stroke:#333333;}#mermaid-svg-xVSYOQeEx7E2tROM .marker.cross{stroke:#333333;}#mermaid-svg-xVSYOQeEx7E2tROM svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-xVSYOQeEx7E2tROM p{margin:0;}#mermaid-svg-xVSYOQeEx7E2tROM .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-xVSYOQeEx7E2tROM .cluster-label text{fill:#333;}#mermaid-svg-xVSYOQeEx7E2tROM .cluster-label span{color:#333;}#mermaid-svg-xVSYOQeEx7E2tROM .cluster-label span p{background-color:transparent;}#mermaid-svg-xVSYOQeEx7E2tROM .label text,#mermaid-svg-xVSYOQeEx7E2tROM span{fill:#333;color:#333;}#mermaid-svg-xVSYOQeEx7E2tROM .node rect,#mermaid-svg-xVSYOQeEx7E2tROM .node circle,#mermaid-svg-xVSYOQeEx7E2tROM .node ellipse,#mermaid-svg-xVSYOQeEx7E2tROM .node polygon,#mermaid-svg-xVSYOQeEx7E2tROM .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-xVSYOQeEx7E2tROM .rough-node .label text,#mermaid-svg-xVSYOQeEx7E2tROM .node .label text,#mermaid-svg-xVSYOQeEx7E2tROM .image-shape .label,#mermaid-svg-xVSYOQeEx7E2tROM .icon-shape .label{text-anchor:middle;}#mermaid-svg-xVSYOQeEx7E2tROM .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-xVSYOQeEx7E2tROM .rough-node .label,#mermaid-svg-xVSYOQeEx7E2tROM .node .label,#mermaid-svg-xVSYOQeEx7E2tROM .image-shape .label,#mermaid-svg-xVSYOQeEx7E2tROM .icon-shape .label{text-align:center;}#mermaid-svg-xVSYOQeEx7E2tROM .node.clickable{cursor:pointer;}#mermaid-svg-xVSYOQeEx7E2tROM .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-xVSYOQeEx7E2tROM .arrowheadPath{fill:#333333;}#mermaid-svg-xVSYOQeEx7E2tROM .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-xVSYOQeEx7E2tROM .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-xVSYOQeEx7E2tROM .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-xVSYOQeEx7E2tROM .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-xVSYOQeEx7E2tROM .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-xVSYOQeEx7E2tROM .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-xVSYOQeEx7E2tROM .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-xVSYOQeEx7E2tROM .cluster text{fill:#333;}#mermaid-svg-xVSYOQeEx7E2tROM .cluster span{color:#333;}#mermaid-svg-xVSYOQeEx7E2tROM div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-xVSYOQeEx7E2tROM .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-xVSYOQeEx7E2tROM rect.text{fill:none;stroke-width:0;}#mermaid-svg-xVSYOQeEx7E2tROM .icon-shape,#mermaid-svg-xVSYOQeEx7E2tROM .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-xVSYOQeEx7E2tROM .icon-shape p,#mermaid-svg-xVSYOQeEx7E2tROM .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-xVSYOQeEx7E2tROM .icon-shape rect,#mermaid-svg-xVSYOQeEx7E2tROM .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-xVSYOQeEx7E2tROM .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-xVSYOQeEx7E2tROM .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-xVSYOQeEx7E2tROM :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u539f\u59cb\u6587\u6863<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u53e5\u5b50\u7d22\u5f15\u7cbe\u51c6\u5339\u914d<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u6458\u8981\u7d22\u5f15\u5168\u5c40\u89c6\u89d2<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u95ee\u9898\u7d22\u5f15\u95ee\u7b54\u5339\u914d<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7236\u6587\u6863\u7d22\u5f15\u4e0a\u4e0b\u6587\u5b8c\u6574<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u591a\u8def\u53ec\u56de<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u91cd\u6392\u5e8f<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u6700\u7ec8\u7ed3\u679c<\/p>\n<p><\/span> <\/p>\n<h4>3.2 \u6df7\u5408\u68c0\u7d22 (Hybrid Search)<\/h4>\n<p>\u6838\u5fc3\u539f\u7406&#xff1a;\u7ed3\u5408\u7a00\u758f\u68c0\u7d22\u548c\u5bc6\u96c6\u68c0\u7d22\u7684\u4f18\u52bf\u3002<\/p>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<table>\n<tr>\u68c0\u7d22\u65b9\u5f0f\u6838\u5fc3\u539f\u7406\u4f18\u52bf\u52a3\u52bf<\/tr>\n<tbody>\n<tr>\n<td>\u7a00\u758f\u68c0\u7d22 (BM25)<\/td>\n<td>\u57fa\u4e8e\u8bcd\u9891\u7edf\u8ba1&#xff0c;\u7cbe\u786e\u5339\u914d\u5173\u952e\u8bcd<\/td>\n<td>\u5173\u952e\u8bcd\u7cbe\u786e\u5339\u914d<\/td>\n<td>\u65e0\u6cd5\u7406\u89e3\u8bed\u4e49<\/td>\n<\/tr>\n<tr>\n<td>\u5bc6\u96c6\u68c0\u7d22 (\u5411\u91cf)<\/td>\n<td>\u57fa\u4e8e\u8bed\u4e49\u76f8\u4f3c\u5ea6&#xff0c;\u7406\u89e3\u542b\u4e49<\/td>\n<td>\u7406\u89e3\u8bed\u4e49\u76f8\u4f3c\u6027<\/td>\n<td>\u9519\u8fc7\u7cbe\u786e\u5173\u952e\u8bcd<\/td>\n<\/tr>\n<tr>\n<td>\u6df7\u5408\u68c0\u7d22<\/td>\n<td>\u878d\u5408\u4e24\u79cd\u7ed3\u679c<\/td>\n<td>\u517c\u987e\u4e24\u8005<\/td>\n<td>\u8ba1\u7b97\u6210\u672c\u8f83\u9ad8<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h5>BM25\u7b97\u6cd5\u539f\u7406<\/h5>\n<p>BM25\u662f\u6700\u6210\u529f\u7684\u7a00\u758f\u68c0\u7d22\u7b97\u6cd5&#xff0c;\u5176\u6838\u5fc3\u601d\u60f3\u662f\u5bf9\u8bcd\u9891\u8fdb\u884c\u9971\u548c\u5904\u7406&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"> <\/p>\n<p>          score <\/p>\n<p>          ( <\/p>\n<p>          D <\/p>\n<p>          , <\/p>\n<p>          Q <\/p>\n<p>          ) <\/p>\n<p>          &#061; <\/p>\n<p>           \u2211 <\/p>\n<p>           i <\/p>\n<p>           n <\/p>\n<p>          IDF <\/p>\n<p>          ( <\/p>\n<p>           q <\/p>\n<p>           i <\/p>\n<p>          ) <\/p>\n<p>          \u00d7 <\/p>\n<p>            f <\/p>\n<p>            ( <\/p>\n<p>             q <\/p>\n<p>             i <\/p>\n<p>            , <\/p>\n<p>            D <\/p>\n<p>            ) <\/p>\n<p>            \u00d7 <\/p>\n<p>            ( <\/p>\n<p>             k <\/p>\n<p>             1 <\/p>\n<p>            &#043; <\/p>\n<p>            1 <\/p>\n<p>            ) <\/p>\n<p>            f <\/p>\n<p>            ( <\/p>\n<p>             q <\/p>\n<p>             i <\/p>\n<p>            , <\/p>\n<p>            D <\/p>\n<p>            ) <\/p>\n<p>            &#043; <\/p>\n<p>             k <\/p>\n<p>             1 <\/p>\n<p>            \u00d7 <\/p>\n<p>            ( <\/p>\n<p>            1 <\/p>\n<p>            \u2212 <\/p>\n<p>            b <\/p>\n<p>            &#043; <\/p>\n<p>            b <\/p>\n<p>            \u00d7 <\/p>\n<p>              \u2223 <\/p>\n<p>              D <\/p>\n<p>              \u2223 <\/p>\n<p>             avgdl <\/p>\n<p>            ) <\/p>\n<p>         \\\\text{score}(D, Q) &#061; \\\\sum_{i}^{n} \\\\text{IDF}(q_i) \\\\times \\\\frac{f(q_i, D) \\\\times (k_1 &#043; 1)}{f(q_i, D) &#043; k_1 \\\\times (1 &#8211; b &#043; b \\\\times \\\\frac{|D|}{\\\\text{avgdl}})} <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord text\"><span class=\"mord\">score<\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">D<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">Q<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.9291em;vertical-align: -1.2777em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.6514em\"><span class=\"\" style=\"top: -1.8723em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">n<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2777em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord text\"><span class=\"mord\">IDF<\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">q<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.8081em;vertical-align: -1.3811em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.427em\"><span class=\"\" style=\"top: -2.11em\"><span class=\"pstrut\" style=\"height: 3.01em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">q<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">D<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0315em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord\">1<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord mathnormal\">b<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord mathnormal\">b<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.01em\"><span class=\"\" style=\"top: -2.655em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord text mtight\"><span class=\"mord mtight\">avgdl<\/span><\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.485em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">\u2223<\/span><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0278em\">D<\/span><span class=\"mord mtight\">\u2223<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.4811em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mclose\">)<\/span><\/span><\/span><span class=\"\" style=\"top: -3.24em\"><span class=\"pstrut\" style=\"height: 3.01em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.687em\"><span class=\"pstrut\" style=\"height: 3.01em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">q<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">D<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0315em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\">1<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3811em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u516c\u5f0f\u89e3\u8bfb&#xff1a;<\/p>\n<ul>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>          f <\/p>\n<p>          ( <\/p>\n<p>           q <\/p>\n<p>           i <\/p>\n<p>          , <\/p>\n<p>          D <\/p>\n<p>          ) <\/p>\n<p>         f(q_i, D) <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">q<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">D<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u8bcd <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"> <\/p>\n<p>           q <\/p>\n<p>           i <\/p>\n<p>         q_i <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">q<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> \u5728\u6587\u6863 <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"> <\/p>\n<p>          D <\/p>\n<p>         D <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">D<\/span><\/span><\/span><\/span><\/span> \u4e2d\u51fa\u73b0\u7684\u6b21\u6570<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>          \u2223 <\/p>\n<p>          D <\/p>\n<p>          \u2223 <\/p>\n<p>         |D| <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord\">\u2223<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">D<\/span><span class=\"mord\">\u2223<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u6587\u6863 <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"> <\/p>\n<p>          D <\/p>\n<p>         D <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">D<\/span><\/span><\/span><\/span><\/span> \u7684\u8bcd\u6570<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>          avgdl <\/p>\n<p>         \\\\text{avgdl} <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8889em;vertical-align: -0.1944em\"><\/span><span class=\"mord text\"><span class=\"mord\">avgdl<\/span><\/span><\/span><\/span><\/span><\/span>&#xff1a;\u6240\u6709\u6587\u6863\u7684\u5e73\u5747\u957f\u5ea6<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>           k <\/p>\n<p>           1 <\/p>\n<p>         k_1 <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8444em;vertical-align: -0.15em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0315em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>&#xff1a;\u9971\u548c\u53c2\u6570&#xff08;\u63a8\u8350 1.2-2.0&#xff09;&#xff0c;\u63a7\u5236\u8bcd\u9891\u7684\u91cd\u8981\u6027<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>          b <\/p>\n<p>         b <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\">b<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u957f\u5ea6\u5f52\u4e00\u5316\u53c2\u6570&#xff08;\u63a8\u8350 0.75&#xff09;&#xff0c;\u63a7\u5236\u6587\u6863\u957f\u5ea6\u7684\u5f71\u54cd<\/li>\n<\/ul>\n<p>BM25\u6539\u8fdb&#xff1a;\u5bf9TF-IDF\u7684\u6539\u8fdb\u5728\u4e8e\u8bcd\u9891\u9971\u548c\u548c\u6587\u6863\u957f\u5ea6\u5f52\u4e00\u5316&#xff0c;\u907f\u514d\u957f\u6587\u6863\u5360\u4f18\u52bf\u3002<\/p>\n<h5>\u5411\u91cf\u68c0\u7d22\u539f\u7406<\/h5>\n<p>Embedding\u6a21\u578b&#xff1a;\u5c06\u6587\u672c\u8f6c\u6362\u4e3a\u9ad8\u7ef4\u5411\u91cf&#xff08;\u901a\u5e38768-3072\u7ef4&#xff09;&#xff0c;\u76f8\u4f3c\u7684\u6587\u672c\u5728\u5411\u91cf\u7a7a\u95f4\u4e2d\u8ddd\u79bb\u66f4\u8fd1\u3002<\/p>\n<p>\u76f8\u4f3c\u5ea6\u8ba1\u7b97&#xff1a;\u901a\u5e38\u4f7f\u7528\u4f59\u5f26\u76f8\u4f3c\u5ea6 <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"> <\/p>\n<p>          similarity <\/p>\n<p>          &#061; <\/p>\n<p>          cos <\/p>\n<p>          \u2061 <\/p>\n<p>          ( <\/p>\n<p>          \u03b8 <\/p>\n<p>          ) <\/p>\n<p>          &#061; <\/p>\n<p>            A <\/p>\n<p>            \u22c5 <\/p>\n<p>            B <\/p>\n<p>            \u2223 <\/p>\n<p>            \u2223 <\/p>\n<p>            A <\/p>\n<p>            \u2223 <\/p>\n<p>            \u2223 <\/p>\n<p>            \u00d7 <\/p>\n<p>            \u2223 <\/p>\n<p>            \u2223 <\/p>\n<p>            B <\/p>\n<p>            \u2223 <\/p>\n<p>            \u2223 <\/p>\n<p>         \\\\text{similarity} &#061; \\\\cos(\\\\theta) &#061; \\\\frac{A \\\\cdot B}{||A|| \\\\times ||B||} <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8889em;vertical-align: -0.1944em\"><\/span><span class=\"mord text\"><span class=\"mord\">similarity<\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mop\">cos<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">\u03b8<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.2963em;vertical-align: -0.936em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3603em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">\u2223\u2223<\/span><span class=\"mord mathnormal\">A<\/span><span class=\"mord\">\u2223\u2223<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\">\u2223\u2223<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0502em\">B<\/span><span class=\"mord\">\u2223\u2223<\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">A<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u22c5<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0502em\">B<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.936em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<h5>RRF\u878d\u5408\u7b97\u6cd5<\/h5>\n<p>RRF (Reciprocal Rank Fusion)&#xff1a;\u5c06\u591a\u4e2a\u68c0\u7d22\u7cfb\u7edf\u7684\u7ed3\u679c\u878d\u5408&#xff0c;\u6838\u5fc3\u516c\u5f0f&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"> <\/p>\n<p>          final_score <\/p>\n<p>          ( <\/p>\n<p>          d <\/p>\n<p>          ) <\/p>\n<p>          &#061; <\/p>\n<p>           \u2211 <\/p>\n<p>            i <\/p>\n<p>            &#061; <\/p>\n<p>            1 <\/p>\n<p>           n <\/p>\n<p>           1 <\/p>\n<p>            k <\/p>\n<p>            &#043; <\/p>\n<p>             rank <\/p>\n<p>             i <\/p>\n<p>            ( <\/p>\n<p>            d <\/p>\n<p>            ) <\/p>\n<p>         \\\\text{final\\\\_score}(d) &#061; \\\\sum_{i&#061;1}^{n} \\\\frac{1}{k &#043; \\\\text{rank}_i(d)} <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.06em;vertical-align: -0.31em\"><\/span><span class=\"mord text\"><span class=\"mord\">final_score<\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">d<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.9291em;vertical-align: -1.2777em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.6514em\"><span class=\"\" style=\"top: -1.8723em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">i<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">n<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2777em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3214em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\"><span class=\"mord text\"><span class=\"mord\">rank<\/span><\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">d<\/span><span class=\"mclose\">)<\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.936em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5176\u4e2d&#xff1a;<\/p>\n<ul>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>          k <\/p>\n<p>         k <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><\/span><\/span><\/span><\/span> \u662f\u5e38\u6570&#xff08;\u901a\u5e38\u53d660&#xff09;&#xff0c;\u63a7\u5236\u6392\u540d\u8870\u51cf\u901f\u5ea6<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>           rank <\/p>\n<p>           i <\/p>\n<p>          ( <\/p>\n<p>          d <\/p>\n<p>          ) <\/p>\n<p>         \\\\text{rank}_i(d) <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord text\"><span class=\"mord\">rank<\/span><\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">d<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span> \u662f\u6587\u6863 <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"> <\/p>\n<p>          d <\/p>\n<p>         d <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\">d<\/span><\/span><\/span><\/span><\/span> \u5728\u7b2c <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"> <\/p>\n<p>          i <\/p>\n<p>         i <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6595em\"><\/span><span class=\"mord mathnormal\">i<\/span><\/span><\/span><\/span><\/span> \u4e2a\u68c0\u7d22\u7cfb\u7edf\u4e2d\u7684\u6392\u540d&#xff08;\u4ece0\u5f00\u59cb&#xff0c;\u6392\u540d\u7b2c1\u7684\u6587\u6863rank&#061;0&#xff09;<\/li>\n<\/ul>\n<p>\u8ba1\u7b97\u793a\u4f8b&#xff1a;<\/p>\n<ul>\n<li>\u6392\u540d\u7b2c1\u7684\u6587\u6863&#xff1a;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>           1 <\/p>\n<p>            60 <\/p>\n<p>            &#043; <\/p>\n<p>            0 <\/p>\n<p>          &#061; <\/p>\n<p>          0.0167 <\/p>\n<p>         \\\\frac{1}{60&#043;0} &#061; 0.0167 <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.2484em;vertical-align: -0.4033em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8451em\"><span class=\"\" style=\"top: -2.655em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">60<\/span><span class=\"mbin mtight\">&#043;<\/span><span class=\"mord mtight\">0<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.394em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.4033em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">0.0167<\/span><\/span><\/span><\/span><\/span><\/li>\n<li>\u6392\u540d\u7b2c2\u7684\u6587\u6863&#xff1a;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>           1 <\/p>\n<p>            60 <\/p>\n<p>            &#043; <\/p>\n<p>            1 <\/p>\n<p>          &#061; <\/p>\n<p>          0.0164 <\/p>\n<p>         \\\\frac{1}{60&#043;1} &#061; 0.0164 <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.2484em;vertical-align: -0.4033em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8451em\"><span class=\"\" style=\"top: -2.655em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">60<\/span><span class=\"mbin mtight\">&#043;<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.394em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.4033em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">0.0164<\/span><\/span><\/span><\/span><\/span><\/li>\n<li>\u6392\u540d\u7b2c10\u7684\u6587\u6863&#xff1a;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>           1 <\/p>\n<p>            60 <\/p>\n<p>            &#043; <\/p>\n<p>            9 <\/p>\n<p>          &#061; <\/p>\n<p>          0.0149 <\/p>\n<p>         \\\\frac{1}{60&#043;9} &#061; 0.0149 <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.2484em;vertical-align: -0.4033em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8451em\"><span class=\"\" style=\"top: -2.655em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">60<\/span><span class=\"mbin mtight\">&#043;<\/span><span class=\"mord mtight\">9<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.394em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.4033em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">0.0149<\/span><\/span><\/span><\/span><\/span><\/li>\n<\/ul>\n<p><span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>retrievers <span class=\"token keyword\">import<\/span> EnsembleRetriever<br \/>\n<span class=\"token keyword\">from<\/span> langchain_community<span class=\"token punctuation\">.<\/span>retrievers <span class=\"token keyword\">import<\/span> BM25Retriever<\/p>\n<p><span class=\"token comment\"># \u7a00\u758f\u68c0\u7d22\u5668 (BM25)<\/span><br \/>\nbm25_retriever <span class=\"token operator\">&#061;<\/span> BM25Retriever<span class=\"token punctuation\">.<\/span>from_documents<span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">,<\/span> k<span class=\"token operator\">&#061;<\/span><span class=\"token number\">5<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u5bc6\u96c6\u68c0\u7d22\u5668 (\u5411\u91cf)<\/span><br \/>\nvector_retriever <span class=\"token operator\">&#061;<\/span> vectorstore<span class=\"token punctuation\">.<\/span>as_retriever<span class=\"token punctuation\">(<\/span>search_kwargs<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;k&#034;<\/span><span class=\"token punctuation\">:<\/span> <span class=\"token number\">5<\/span><span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u878d\u5408\u68c0\u7d22\u5668 &#8211; \u4f7f\u7528RRF\u7b97\u6cd5<\/span><br \/>\nensemble_retriever <span class=\"token operator\">&#061;<\/span> EnsembleRetriever<span class=\"token punctuation\">(<\/span><br \/>\n    retrievers<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span>bm25_retriever<span class=\"token punctuation\">,<\/span> vector_retriever<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    weights<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token number\">0.5<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">0.5<\/span><span class=\"token punctuation\">]<\/span>  <span class=\"token comment\"># \u53ef\u8c03\u6574\u6743\u91cd<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p>results <span class=\"token operator\">&#061;<\/span> ensemble_retriever<span class=\"token punctuation\">.<\/span>get_relevant_documents<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#034;\u4ec0\u4e48\u662fRAG&#xff1f;&#034;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<h4>3.3 \u53e5\u5b50\u7a97\u53e3\u68c0\u7d22 (Sentence Window Retriever)<\/h4>\n<p>\u6838\u5fc3\u6982\u5ff5&#xff1a;\u53e5\u5b50\u7a97\u53e3\u68c0\u7d22\u662f\u4e00\u79cd\u5e73\u8861\u7cbe\u5ea6\u548c\u4e0a\u4e0b\u6587\u7684\u6280\u672f\u3002<\/p>\n<p>\u5de5\u4f5c\u539f\u7406&#xff1a;<\/p>\n<li>\u5c0f\u7a97\u53e3\u68c0\u7d22&#xff1a;\u5c06\u6587\u6863\u5206\u5272\u6210\u5355\u53e5&#xff0c;\u7528\u5355\u53e5\u8fdb\u884c\u5411\u91cf\u68c0\u7d22<\/li>\n<li>\u5927\u7a97\u53e3\u8fd4\u56de&#xff1a;\u68c0\u7d22\u5230\u76f8\u5173\u53e5\u5b50\u540e&#xff0c;\u8fd4\u56de\u5176\u524d\u540e\u82e5\u5e72\u53e5\u7ec4\u6210\u7684\u7a97\u53e3<\/li>\n<p>\u4f18\u52bf&#xff1a;<\/p>\n<ul>\n<li>\u68c0\u7d22\u65f6\u7528\u5c0f\u7a97\u53e3&#xff0c;\u7cbe\u5ea6\u9ad8&#xff08;\u5411\u91cf\u66f4\u7cbe\u51c6&#xff09;<\/li>\n<li>\u8fd4\u56de\u65f6\u7528\u5927\u7a97\u53e3&#xff0c;\u4e0a\u4e0b\u6587\u5b8c\u6574<\/li>\n<\/ul>\n<p><span class=\"token keyword\">from<\/span> llama_index<span class=\"token punctuation\">.<\/span>node_parser <span class=\"token keyword\">import<\/span> SentenceWindowNodeParser<br \/>\n<span class=\"token keyword\">from<\/span> llama_index <span class=\"token keyword\">import<\/span> VectorStoreIndex<span class=\"token punctuation\">,<\/span> Document<br \/>\n<span class=\"token keyword\">from<\/span> llama_index<span class=\"token punctuation\">.<\/span>postprocessor <span class=\"token keyword\">import<\/span> MetadataReplacementPostprocessor<\/p>\n<p><span class=\"token comment\"># \u521b\u5efa\u53e5\u5b50\u7a97\u53e3\u89e3\u6790\u5668<\/span><br \/>\nnode_parser <span class=\"token operator\">&#061;<\/span> SentenceWindowNodeParser<span class=\"token punctuation\">.<\/span>from_defaults<span class=\"token punctuation\">(<\/span><br \/>\n    window_size<span class=\"token operator\">&#061;<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u7a97\u53e3\u5927\u5c0f&#xff1a;\u524d\u540e\u54043\u53e5<\/span><br \/>\n    window_metadata_key<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;window&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    original_text_metadata_key<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;original_text&#034;<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u89e3\u6790\u6587\u6863\u4e3a\u8282\u70b9<\/span><br \/>\nnodes <span class=\"token operator\">&#061;<\/span> node_parser<span class=\"token punctuation\">.<\/span>get_nodes_from_documents<span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u521b\u5efa\u7d22\u5f15<\/span><br \/>\nindex <span class=\"token operator\">&#061;<\/span> VectorStoreIndex<span class=\"token punctuation\">(<\/span>nodes<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u521b\u5efa\u67e5\u8be2\u5f15\u64ce&#xff08;\u5e26\u7a97\u53e3\u6062\u590d&#xff09;<\/span><br \/>\nquery_engine <span class=\"token operator\">&#061;<\/span> index<span class=\"token punctuation\">.<\/span>as_query_engine<span class=\"token punctuation\">(<\/span><br \/>\n    similarity_top_k<span class=\"token operator\">&#061;<\/span><span class=\"token number\">2<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    node_postprocessors<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><br \/>\n        MetadataReplacementPostprocessor<span class=\"token punctuation\">(<\/span>target_metadata_key<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;window&#034;<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token punctuation\">]<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p>response <span class=\"token operator\">&#061;<\/span> query_engine<span class=\"token punctuation\">.<\/span>query<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#034;RAG\u6709\u54ea\u4e9b\u4f18\u5316\u65b9\u5f0f&#xff1f;&#034;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<h4>3.4 \u7236\u5b50\u6587\u6863\u68c0\u7d22 (Parent Document Retriever)<\/h4>\n<p>\u6838\u5fc3\u6982\u5ff5&#xff1a;\u7528\u5c0f\u5757\u68c0\u7d22&#xff0c;\u8fd4\u56de\u5927\u5757\u3002<\/p>\n<p>\u8bbe\u8ba1\u601d\u60f3<\/p>\n<p>   #mermaid-svg-eqBdhxYmxw3HeyM6{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-eqBdhxYmxw3HeyM6 .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-eqBdhxYmxw3HeyM6 .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-eqBdhxYmxw3HeyM6 .error-icon{fill:#552222;}#mermaid-svg-eqBdhxYmxw3HeyM6 .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-eqBdhxYmxw3HeyM6 .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-eqBdhxYmxw3HeyM6 .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-eqBdhxYmxw3HeyM6 .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-eqBdhxYmxw3HeyM6 .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-eqBdhxYmxw3HeyM6 .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-eqBdhxYmxw3HeyM6 .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-eqBdhxYmxw3HeyM6 .marker{fill:#333333;stroke:#333333;}#mermaid-svg-eqBdhxYmxw3HeyM6 .marker.cross{stroke:#333333;}#mermaid-svg-eqBdhxYmxw3HeyM6 svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-eqBdhxYmxw3HeyM6 p{margin:0;}#mermaid-svg-eqBdhxYmxw3HeyM6 .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-eqBdhxYmxw3HeyM6 .cluster-label text{fill:#333;}#mermaid-svg-eqBdhxYmxw3HeyM6 .cluster-label span{color:#333;}#mermaid-svg-eqBdhxYmxw3HeyM6 .cluster-label span p{background-color:transparent;}#mermaid-svg-eqBdhxYmxw3HeyM6 .label text,#mermaid-svg-eqBdhxYmxw3HeyM6 span{fill:#333;color:#333;}#mermaid-svg-eqBdhxYmxw3HeyM6 .node rect,#mermaid-svg-eqBdhxYmxw3HeyM6 .node circle,#mermaid-svg-eqBdhxYmxw3HeyM6 .node ellipse,#mermaid-svg-eqBdhxYmxw3HeyM6 .node polygon,#mermaid-svg-eqBdhxYmxw3HeyM6 .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-eqBdhxYmxw3HeyM6 .rough-node .label text,#mermaid-svg-eqBdhxYmxw3HeyM6 .node .label text,#mermaid-svg-eqBdhxYmxw3HeyM6 .image-shape .label,#mermaid-svg-eqBdhxYmxw3HeyM6 .icon-shape .label{text-anchor:middle;}#mermaid-svg-eqBdhxYmxw3HeyM6 .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-eqBdhxYmxw3HeyM6 .rough-node .label,#mermaid-svg-eqBdhxYmxw3HeyM6 .node .label,#mermaid-svg-eqBdhxYmxw3HeyM6 .image-shape .label,#mermaid-svg-eqBdhxYmxw3HeyM6 .icon-shape .label{text-align:center;}#mermaid-svg-eqBdhxYmxw3HeyM6 .node.clickable{cursor:pointer;}#mermaid-svg-eqBdhxYmxw3HeyM6 .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-eqBdhxYmxw3HeyM6 .arrowheadPath{fill:#333333;}#mermaid-svg-eqBdhxYmxw3HeyM6 .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-eqBdhxYmxw3HeyM6 .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-eqBdhxYmxw3HeyM6 .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-eqBdhxYmxw3HeyM6 .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-eqBdhxYmxw3HeyM6 .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-eqBdhxYmxw3HeyM6 .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-eqBdhxYmxw3HeyM6 .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-eqBdhxYmxw3HeyM6 .cluster text{fill:#333;}#mermaid-svg-eqBdhxYmxw3HeyM6 .cluster span{color:#333;}#mermaid-svg-eqBdhxYmxw3HeyM6 div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-eqBdhxYmxw3HeyM6 .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-eqBdhxYmxw3HeyM6 rect.text{fill:none;stroke-width:0;}#mermaid-svg-eqBdhxYmxw3HeyM6 .icon-shape,#mermaid-svg-eqBdhxYmxw3HeyM6 .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-eqBdhxYmxw3HeyM6 .icon-shape p,#mermaid-svg-eqBdhxYmxw3HeyM6 .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-eqBdhxYmxw3HeyM6 .icon-shape rect,#mermaid-svg-eqBdhxYmxw3HeyM6 .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-eqBdhxYmxw3HeyM6 .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-eqBdhxYmxw3HeyM6 .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-eqBdhxYmxw3HeyM6 :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>          <span class=\"edgeLabel\"><\/span> <\/p>\n<p>          <span class=\"edgeLabel\"><\/span> <\/p>\n<p>          <span class=\"edgeLabel\"><\/span> <\/p>\n<p>          <span class=\"edgeLabel\"><\/span> <\/p>\n<p>          <span class=\"edgeLabel\"><\/span> <\/p>\n<p>          <span class=\"edgeLabel\"><\/span> <\/p>\n<p>          <span class=\"edgeLabel\"><\/span> <\/p>\n<p>          <span class=\"edgeLabel\"><\/span> <\/p>\n<p>          <span class=\"edgeLabel\"><\/span> <\/p>\n<p>          <span class=\"nodeLabel\"><\/p>\n<p>\u5927\u6587\u6863 1000\u5b57<\/p>\n<p><\/span> <\/p>\n<p>          <span class=\"nodeLabel\"><\/p>\n<p>\u5206\u5272<\/p>\n<p><\/span> <\/p>\n<p>          <span class=\"nodeLabel\"><\/p>\n<p>\u5c0f\u6587\u68631200\u5b57<\/p>\n<p><\/span> <\/p>\n<p>          <span class=\"nodeLabel\"><\/p>\n<p>\u5c0f\u6587\u68632200\u5b57<\/p>\n<p><\/span> <\/p>\n<p>          <span class=\"nodeLabel\"><\/p>\n<p>\u5c0f\u6587\u68633200\u5b57<\/p>\n<p><\/span> <\/p>\n<p>          <span class=\"nodeLabel\"><\/p>\n<p>\u5411\u91cf\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>          <span class=\"nodeLabel\"><\/p>\n<p>\u5339\u914d\u5230\u5c0f\u6587\u68632<\/p>\n<p><\/span> <\/p>\n<p>          <span class=\"nodeLabel\"><\/p>\n<p>\u8fd4\u56de\u7236\u6587\u6863\u5b8c\u65741000\u5b57<\/p>\n<p><\/span> <\/p>\n<p>\u5de5\u4f5c\u539f\u7406&#xff1a;<\/p>\n<li>\u5c06\u5927\u6587\u6863\u5206\u5272\u6210\u7236\u5b50\u4e24\u4e2a\u5c42\u6b21<\/li>\n<li>\u7528\u5c0f\u6587\u6863\u8fdb\u884c\u5411\u91cf\u68c0\u7d22&#xff08;\u7cbe\u5ea6\u9ad8&#xff09;<\/li>\n<li>\u8fd4\u56de\u5bf9\u5e94\u7684\u5927\u6587\u6863&#xff08;\u4e0a\u4e0b\u6587\u5b8c\u6574&#xff09;<\/li>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u7236\u6587\u6863&#xff1a;\u8f83\u5927\u7684\u6587\u6863\u5757&#xff08;\u59821000\u5b57&#xff09;&#xff0c;\u5305\u542b\u5b8c\u6574\u7684\u4e0a\u4e0b\u6587<\/li>\n<li>\u5b50\u6587\u6863&#xff1a;\u8f83\u5c0f\u7684\u6587\u6863\u5757&#xff08;\u5982200\u5b57&#xff09;&#xff0c;\u7528\u4e8e\u7cbe\u51c6\u68c0\u7d22<\/li>\n<\/ul>\n<p><span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>retrievers <span class=\"token keyword\">import<\/span> ParentDocumentRetriever<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>storage <span class=\"token keyword\">import<\/span> InMemoryStore<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>text_splitter <span class=\"token keyword\">import<\/span> RecursiveCharacterTextSplitter<\/p>\n<p><span class=\"token comment\"># \u5c0f\u6587\u6863\u5206\u5272\u5668&#xff08;\u7528\u4e8e\u68c0\u7d22&#xff09;<\/span><br \/>\nchild_splitter <span class=\"token operator\">&#061;<\/span> RecursiveCharacterTextSplitter<span class=\"token punctuation\">(<\/span><br \/>\n    chunk_size<span class=\"token operator\">&#061;<\/span><span class=\"token number\">200<\/span><span class=\"token punctuation\">,<\/span>      <span class=\"token comment\"># 200\u5b57\u7b26\u4e00\u4e2a\u5757<\/span><br \/>\n    chunk_overlap<span class=\"token operator\">&#061;<\/span><span class=\"token number\">50<\/span><span class=\"token punctuation\">,<\/span>    <span class=\"token comment\"># \u91cd\u53e050\u5b57\u7b26<\/span><br \/>\n    separators<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;\\\\n\\\\n&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034;\\\\n&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034;\u3002&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034;&#xff01;&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034;&#xff1f;&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034;&#xff0c;&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034; &#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034;&#034;<\/span><span class=\"token punctuation\">]<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u5927\u6587\u6863\u5206\u5272\u5668&#xff08;\u7528\u4e8e\u8fd4\u56de&#xff09;<\/span><br \/>\nparent_splitter <span class=\"token operator\">&#061;<\/span> RecursiveCharacterTextSplitter<span class=\"token punctuation\">(<\/span><br \/>\n    chunk_size<span class=\"token operator\">&#061;<\/span><span class=\"token number\">1000<\/span><span class=\"token punctuation\">,<\/span>     <span class=\"token comment\"># 1000\u5b57\u7b26\u4e00\u4e2a\u5757<\/span><br \/>\n    chunk_overlap<span class=\"token operator\">&#061;<\/span><span class=\"token number\">100<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    separators<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;\\\\n\\\\n&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034;\\\\n&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034;\u3002&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034;&#xff01;&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token string\">&#034;&#xff1f;&#034;<\/span><span class=\"token punctuation\">]<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u521b\u5efa\u7236\u6587\u6863\u68c0\u7d22\u5668<\/span><br \/>\nretriever <span class=\"token operator\">&#061;<\/span> ParentDocumentRetriever<span class=\"token punctuation\">(<\/span><br \/>\n    vectorstore<span class=\"token operator\">&#061;<\/span>vectorstore<span class=\"token punctuation\">,<\/span><br \/>\n    docstore<span class=\"token operator\">&#061;<\/span>store<span class=\"token punctuation\">,<\/span><br \/>\n    child_splitter<span class=\"token operator\">&#061;<\/span>child_splitter<span class=\"token punctuation\">,<\/span><br \/>\n    parent_splitter<span class=\"token operator\">&#061;<\/span>parent_splitter<br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<h4>3.5 \u6458\u8981\u68c0\u7d22 (Summary Retriever)<\/h4>\n<p>  #mermaid-svg-dsknaoLZTw9N2qX6{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-dsknaoLZTw9N2qX6 .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-dsknaoLZTw9N2qX6 .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-dsknaoLZTw9N2qX6 .error-icon{fill:#552222;}#mermaid-svg-dsknaoLZTw9N2qX6 .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-dsknaoLZTw9N2qX6 .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-dsknaoLZTw9N2qX6 .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-dsknaoLZTw9N2qX6 .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-dsknaoLZTw9N2qX6 .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-dsknaoLZTw9N2qX6 .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-dsknaoLZTw9N2qX6 .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-dsknaoLZTw9N2qX6 .marker{fill:#333333;stroke:#333333;}#mermaid-svg-dsknaoLZTw9N2qX6 .marker.cross{stroke:#333333;}#mermaid-svg-dsknaoLZTw9N2qX6 svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-dsknaoLZTw9N2qX6 p{margin:0;}#mermaid-svg-dsknaoLZTw9N2qX6 .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-dsknaoLZTw9N2qX6 .cluster-label text{fill:#333;}#mermaid-svg-dsknaoLZTw9N2qX6 .cluster-label span{color:#333;}#mermaid-svg-dsknaoLZTw9N2qX6 .cluster-label span p{background-color:transparent;}#mermaid-svg-dsknaoLZTw9N2qX6 .label text,#mermaid-svg-dsknaoLZTw9N2qX6 span{fill:#333;color:#333;}#mermaid-svg-dsknaoLZTw9N2qX6 .node rect,#mermaid-svg-dsknaoLZTw9N2qX6 .node circle,#mermaid-svg-dsknaoLZTw9N2qX6 .node ellipse,#mermaid-svg-dsknaoLZTw9N2qX6 .node polygon,#mermaid-svg-dsknaoLZTw9N2qX6 .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-dsknaoLZTw9N2qX6 .rough-node .label text,#mermaid-svg-dsknaoLZTw9N2qX6 .node .label text,#mermaid-svg-dsknaoLZTw9N2qX6 .image-shape .label,#mermaid-svg-dsknaoLZTw9N2qX6 .icon-shape .label{text-anchor:middle;}#mermaid-svg-dsknaoLZTw9N2qX6 .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-dsknaoLZTw9N2qX6 .rough-node .label,#mermaid-svg-dsknaoLZTw9N2qX6 .node .label,#mermaid-svg-dsknaoLZTw9N2qX6 .image-shape .label,#mermaid-svg-dsknaoLZTw9N2qX6 .icon-shape .label{text-align:center;}#mermaid-svg-dsknaoLZTw9N2qX6 .node.clickable{cursor:pointer;}#mermaid-svg-dsknaoLZTw9N2qX6 .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-dsknaoLZTw9N2qX6 .arrowheadPath{fill:#333333;}#mermaid-svg-dsknaoLZTw9N2qX6 .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-dsknaoLZTw9N2qX6 .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-dsknaoLZTw9N2qX6 .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-dsknaoLZTw9N2qX6 .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-dsknaoLZTw9N2qX6 .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-dsknaoLZTw9N2qX6 .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-dsknaoLZTw9N2qX6 .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-dsknaoLZTw9N2qX6 .cluster text{fill:#333;}#mermaid-svg-dsknaoLZTw9N2qX6 .cluster span{color:#333;}#mermaid-svg-dsknaoLZTw9N2qX6 div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-dsknaoLZTw9N2qX6 .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-dsknaoLZTw9N2qX6 rect.text{fill:none;stroke-width:0;}#mermaid-svg-dsknaoLZTw9N2qX6 .icon-shape,#mermaid-svg-dsknaoLZTw9N2qX6 .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-dsknaoLZTw9N2qX6 .icon-shape p,#mermaid-svg-dsknaoLZTw9N2qX6 .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-dsknaoLZTw9N2qX6 .icon-shape rect,#mermaid-svg-dsknaoLZTw9N2qX6 .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-dsknaoLZTw9N2qX6 .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-dsknaoLZTw9N2qX6 .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-dsknaoLZTw9N2qX6 :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u539f\u59cb\u6587\u68635000\u5b57<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u751f\u6210\u6458\u8981200\u5b57<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u6458\u8981\u7d22\u5f15\u5411\u91cf\u5b58\u50a8<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7528\u6237\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5411\u91cf\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5339\u914d\u6458\u8981<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u8fd4\u56de\u5b8c\u6574\u6587\u6863<\/p>\n<p><\/span> <\/p>\n<p>\u6838\u5fc3\u6982\u5ff5&#xff1a;\u7528\u6587\u6863\u6458\u8981\u8fdb\u884c\u68c0\u7d22&#xff0c;\u8fd4\u56de\u5b8c\u6574\u6587\u6863\u3002<\/p>\n<p>\u4e3a\u4ec0\u4e48\u9700\u8981\u6458\u8981\u68c0\u7d22&#xff1f;<\/p>\n<p>\u573a\u666f&#xff1a;\u7528\u6237\u95ee&#034;RAG\u6280\u672f\u5728\u4f01\u4e1a\u4e2d\u7684\u5e94\u7528&#034;<\/p>\n<p>\u95ee\u9898&#xff1a;<\/p>\n<ul>\n<li>\u539f\u59cb\u6587\u6863\u592a\u957f&#xff08;5000\u5b57&#xff09;&#xff0c;\u5411\u91cf\u68c0\u7d22\u5bb9\u6613\u805a\u7126\u7ec6\u8282\u800c\u5ffd\u7565\u4e3b\u9898<\/li>\n<li>\u6458\u8981\u6982\u62ec\u4e86\u6587\u6863\u7684\u6838\u5fc3\u5185\u5bb9&#xff08;200\u5b57&#xff09;<\/li>\n<\/ul>\n<p>\u89e3\u51b3\u65b9\u6848&#xff1a;<\/p>\n<ul>\n<li>\u7528\u6458\u8981\u5411\u91cf\u8fdb\u884c\u68c0\u7d22&#xff08;\u5339\u914d\u4e3b\u9898&#xff09;<\/li>\n<li>\u8fd4\u56de\u5b8c\u6574\u6587\u6863&#xff08;\u63d0\u4f9b\u7ec6\u8282&#xff09;<\/li>\n<\/ul>\n<p>\u5de5\u4f5c\u539f\u7406&#xff1a;<\/p>\n<li>\u4e3a\u6bcf\u4e2a\u6587\u6863\u751f\u6210\u6458\u8981<\/li>\n<li>\u7528\u6458\u8981\u6784\u5efa\u5411\u91cf\u7d22\u5f15<\/li>\n<li>\u68c0\u7d22\u65f6\u5339\u914d\u6458\u8981&#xff0c;\u8fd4\u56de\u5b8c\u6574\u6587\u6863<\/li>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u6458\u8981\u7d22\u5f15&#xff1a;\u5b58\u50a8\u6587\u6863\u6458\u8981\u7684\u5411\u91cf\u7d22\u5f15&#xff0c;\u7528\u4e8e\u4e3b\u9898\u7ea7\u522b\u7684\u68c0\u7d22<\/li>\n<li>\u5168\u5c40\u89c6\u89d2&#xff1a;\u6458\u8981\u63d0\u4f9b\u6587\u6863\u7684\u5168\u5c40\u89c6\u56fe&#xff0c;\u907f\u514d\u9677\u5165\u7ec6\u8282<\/li>\n<\/ul>\n<p><span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>prompts <span class=\"token keyword\">import<\/span> PromptTemplate<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>llms <span class=\"token keyword\">import<\/span> OpenAI<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>schema <span class=\"token keyword\">import<\/span> Document<\/p>\n<p><span class=\"token comment\"># 1. \u6458\u8981\u751f\u6210Prompt<\/span><br \/>\nsummary_prompt <span class=\"token operator\">&#061;<\/span> PromptTemplate<span class=\"token punctuation\">(<\/span><br \/>\n    template<span class=\"token operator\">&#061;<\/span><span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u8bf7\u4e3a\u4ee5\u4e0b\u6587\u6863\u751f\u6210\u4e00\u4e2a\u7b80\u6d01\u7684\u6458\u8981&#xff08;100\u5b57\u4ee5\u5185&#xff09;&#xff1a;<\/p>\n<p>\u6587\u6863:<br \/>\n{context}<\/p>\n<p>\u6458\u8981:&#034;&#034;&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    input_variables<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;context&#034;<\/span><span class=\"token punctuation\">]<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># 2. \u4e3a\u6587\u6863\u751f\u6210\u6458\u8981<\/span><br \/>\n<span class=\"token keyword\">def<\/span> <span class=\"token function\">generate_summary<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    llm <span class=\"token operator\">&#061;<\/span> OpenAI<span class=\"token punctuation\">(<\/span>temperature<span class=\"token operator\">&#061;<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    summaries <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p>    <span class=\"token keyword\">for<\/span> i<span class=\"token punctuation\">,<\/span> doc <span class=\"token keyword\">in<\/span> <span class=\"token builtin\">enumerate<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token keyword\">print<\/span><span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#034;\u5904\u7406\u7b2c <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>i<span class=\"token operator\">&#043;<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">\/<\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span><span class=\"token builtin\">len<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\"> \u4e2a\u6587\u6863&#8230;&#034;<\/span><\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u751f\u6210\u6458\u8981<\/span><br \/>\n        summary <span class=\"token operator\">&#061;<\/span> llm<span class=\"token punctuation\">(<\/span>summary_prompt<span class=\"token punctuation\">.<\/span><span class=\"token builtin\">format<\/span><span class=\"token punctuation\">(<\/span>context<span class=\"token operator\">&#061;<\/span>doc<span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>        summaries<span class=\"token punctuation\">.<\/span>append<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">{<\/span><br \/>\n            <span class=\"token string\">&#034;summary&#034;<\/span><span class=\"token punctuation\">:<\/span> summary<span class=\"token punctuation\">,<\/span><br \/>\n            <span class=\"token string\">&#034;original_doc&#034;<\/span><span class=\"token punctuation\">:<\/span> doc<br \/>\n        <span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token keyword\">return<\/span> summaries<\/p>\n<p><span class=\"token comment\"># 3. \u4e3a\u77e5\u8bc6\u5e93\u751f\u6210\u6458\u8981<\/span><br \/>\nsummaries <span class=\"token operator\">&#061;<\/span> generate_summary<span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># 4. \u7528\u6458\u8981\u6784\u5efa\u7d22\u5f15<\/span><br \/>\nsummary_docs <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><br \/>\n    Document<span class=\"token punctuation\">(<\/span><br \/>\n        page_content<span class=\"token operator\">&#061;<\/span>s<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;summary&#034;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        metadata<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;doc_id&#034;<\/span><span class=\"token punctuation\">:<\/span> i<span class=\"token punctuation\">}<\/span><br \/>\n    <span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">for<\/span> i<span class=\"token punctuation\">,<\/span> s <span class=\"token keyword\">in<\/span> <span class=\"token builtin\">enumerate<\/span><span class=\"token punctuation\">(<\/span>summaries<span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token punctuation\">]<\/span><\/p>\n<p>summary_index <span class=\"token operator\">&#061;<\/span> Chroma<span class=\"token punctuation\">.<\/span>from_documents<span class=\"token punctuation\">(<\/span><br \/>\n    summary_docs<span class=\"token punctuation\">,<\/span><br \/>\n    embeddings<span class=\"token punctuation\">,<\/span><br \/>\n    collection_name<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;summary_index&#034;<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># 5. \u68c0\u7d22\u51fd\u6570<\/span><br \/>\n<span class=\"token keyword\">def<\/span> <span class=\"token function\">retrieve_by_summary<\/span><span class=\"token punctuation\">(<\/span>query<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u901a\u8fc7\u6458\u8981\u68c0\u7d22&#xff0c;\u8fd4\u56de\u539f\u59cb\u6587\u6863&#034;&#034;&#034;<\/span><br \/>\n    <span class=\"token comment\"># \u627e\u5230\u76f8\u4f3c\u7684\u6458\u8981<\/span><br \/>\n    similar_summaries <span class=\"token operator\">&#061;<\/span> summary_index<span class=\"token punctuation\">.<\/span>similarity_search<span class=\"token punctuation\">(<\/span>query<span class=\"token punctuation\">,<\/span> k<span class=\"token operator\">&#061;<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token keyword\">if<\/span> <span class=\"token keyword\">not<\/span> similar_summaries<span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token keyword\">return<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p>    <span class=\"token comment\"># \u83b7\u53d6\u5bf9\u5e94\u7684\u539f\u59cb\u6587\u6863<\/span><br \/>\n    doc_id <span class=\"token operator\">&#061;<\/span> similar_summaries<span class=\"token punctuation\">[<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>metadata<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;doc_id&#034;<\/span><span class=\"token punctuation\">]<\/span><br \/>\n    <span class=\"token keyword\">return<\/span> <span class=\"token punctuation\">[<\/span>summaries<span class=\"token punctuation\">[<\/span>doc_id<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;original_doc&#034;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p><span class=\"token comment\"># \u4f7f\u7528\u793a\u4f8b<\/span><br \/>\nresults <span class=\"token operator\">&#061;<\/span> retrieve_by_summary<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#034;RAG\u5728\u4f01\u4e1a\u4e2d\u7684\u5e94\u7528\u573a\u666f&#034;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<h4>3.6 \u5047\u8bbe\u6027\u95ee\u9898\u68c0\u7d22 (Hypothetical Questions Retriever)<\/h4>\n<p>  #mermaid-svg-VhHYhUZxffWjKkhh{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-VhHYhUZxffWjKkhh .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-VhHYhUZxffWjKkhh .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-VhHYhUZxffWjKkhh .error-icon{fill:#552222;}#mermaid-svg-VhHYhUZxffWjKkhh .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-VhHYhUZxffWjKkhh .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-VhHYhUZxffWjKkhh .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-VhHYhUZxffWjKkhh .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-VhHYhUZxffWjKkhh .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-VhHYhUZxffWjKkhh .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-VhHYhUZxffWjKkhh .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-VhHYhUZxffWjKkhh .marker{fill:#333333;stroke:#333333;}#mermaid-svg-VhHYhUZxffWjKkhh .marker.cross{stroke:#333333;}#mermaid-svg-VhHYhUZxffWjKkhh svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-VhHYhUZxffWjKkhh p{margin:0;}#mermaid-svg-VhHYhUZxffWjKkhh .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-VhHYhUZxffWjKkhh .cluster-label text{fill:#333;}#mermaid-svg-VhHYhUZxffWjKkhh .cluster-label span{color:#333;}#mermaid-svg-VhHYhUZxffWjKkhh .cluster-label span p{background-color:transparent;}#mermaid-svg-VhHYhUZxffWjKkhh .label text,#mermaid-svg-VhHYhUZxffWjKkhh span{fill:#333;color:#333;}#mermaid-svg-VhHYhUZxffWjKkhh .node rect,#mermaid-svg-VhHYhUZxffWjKkhh .node circle,#mermaid-svg-VhHYhUZxffWjKkhh .node ellipse,#mermaid-svg-VhHYhUZxffWjKkhh .node polygon,#mermaid-svg-VhHYhUZxffWjKkhh .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-VhHYhUZxffWjKkhh .rough-node .label text,#mermaid-svg-VhHYhUZxffWjKkhh .node .label text,#mermaid-svg-VhHYhUZxffWjKkhh .image-shape .label,#mermaid-svg-VhHYhUZxffWjKkhh .icon-shape .label{text-anchor:middle;}#mermaid-svg-VhHYhUZxffWjKkhh .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-VhHYhUZxffWjKkhh .rough-node .label,#mermaid-svg-VhHYhUZxffWjKkhh .node .label,#mermaid-svg-VhHYhUZxffWjKkhh .image-shape .label,#mermaid-svg-VhHYhUZxffWjKkhh .icon-shape .label{text-align:center;}#mermaid-svg-VhHYhUZxffWjKkhh .node.clickable{cursor:pointer;}#mermaid-svg-VhHYhUZxffWjKkhh .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-VhHYhUZxffWjKkhh .arrowheadPath{fill:#333333;}#mermaid-svg-VhHYhUZxffWjKkhh .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-VhHYhUZxffWjKkhh .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-VhHYhUZxffWjKkhh .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-VhHYhUZxffWjKkhh .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-VhHYhUZxffWjKkhh .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-VhHYhUZxffWjKkhh .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-VhHYhUZxffWjKkhh .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-VhHYhUZxffWjKkhh .cluster text{fill:#333;}#mermaid-svg-VhHYhUZxffWjKkhh .cluster span{color:#333;}#mermaid-svg-VhHYhUZxffWjKkhh div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-VhHYhUZxffWjKkhh .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-VhHYhUZxffWjKkhh rect.text{fill:none;stroke-width:0;}#mermaid-svg-VhHYhUZxffWjKkhh .icon-shape,#mermaid-svg-VhHYhUZxffWjKkhh .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-VhHYhUZxffWjKkhh .icon-shape p,#mermaid-svg-VhHYhUZxffWjKkhh .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-VhHYhUZxffWjKkhh .icon-shape rect,#mermaid-svg-VhHYhUZxffWjKkhh .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-VhHYhUZxffWjKkhh .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-VhHYhUZxffWjKkhh .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-VhHYhUZxffWjKkhh :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u539f\u59cb\u6587\u6863<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>LLM\u751f\u6210\u5047\u8bbe\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u95ee\u98981Python\u662f\u4ec0\u4e48\u65f6\u5019\u521b\u5efa\u7684?<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u95ee\u98982Python\u7684\u521b\u59cb\u4eba\u662f\u8c01?<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u95ee\u98983Python\u6709\u4ec0\u4e48\u7279\u70b9?<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u95ee\u9898\u7d22\u5f15\u5411\u91cf\u5b58\u50a8<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7528\u6237\u95ee\u9898Python\u7684\u5386\u53f2<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5411\u91cf\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5339\u914d\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u8fd4\u56de\u539f\u59cb\u6587\u6863<\/p>\n<p><\/span> <\/p>\n<p>\u6838\u5fc3\u6982\u5ff5&#xff1a;\u9884\u5148\u751f\u6210\u6587\u6863\u53ef\u80fd\u56de\u7b54\u7684\u95ee\u9898&#xff0c;\u7528\u8fd9\u4e9b\u95ee\u9898\u8fdb\u884c\u68c0\u7d22\u3002<\/p>\n<p>\u5e94\u7528\u573a\u666f<\/p>\n<p>\u6587\u6863&#xff1a;\u201cPython\u662f\u4e00\u95e8\u9ad8\u7ea7\u7f16\u7a0b\u8bed\u8a00&#xff0c;\u7531Guido van Rossum\u4e8e1991\u5e74\u521b\u5efa\u2026\u201d<\/p>\n<p>\u751f\u6210\u5047\u8bbe\u6027\u95ee\u9898&#xff1a;<\/p>\n<ul>\n<li>\u201cPython\u662f\u4ec0\u4e48\u65f6\u5019\u521b\u5efa\u7684&#xff1f;\u201d<\/li>\n<li>\u201cPython\u7684\u521b\u59cb\u4eba\u662f\u8c01&#xff1f;\u201d<\/li>\n<li>\u201cPython\u6709\u4ec0\u4e48\u7279\u70b9&#xff1f;\u201d<\/li>\n<\/ul>\n<p>\u68c0\u7d22\u65f6&#xff1a;\u7528\u6237\u95ee&#034;Python\u7684\u5386\u53f2&#034; \u2192 \u5339\u914d\u5230&#034;Python\u662f\u4ec0\u4e48\u65f6\u5019\u521b\u5efa\u7684&#xff1f;&#034; \u2192 \u8fd4\u56de\u539f\u6587\u6863<\/p>\n<p>\u5de5\u4f5c\u539f\u7406&#xff1a;<\/p>\n<li>\u4e3a\u6bcf\u4e2a\u6587\u6863\u9884\u5148\u751f\u62103-5\u4e2a\u53ef\u80fd\u7684\u95ee\u9898<\/li>\n<li>\u7528\u8fd9\u4e9b\u95ee\u9898\u6784\u5efa\u5411\u91cf\u7d22\u5f15<\/li>\n<li>\u7528\u6237\u67e5\u8be2\u65f6&#xff0c;\u5339\u914d\u8fd9\u4e9b\u95ee\u9898&#xff0c;\u8fd4\u56de\u5bf9\u5e94\u6587\u6863<\/li>\n<p><span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>prompts <span class=\"token keyword\">import<\/span> PromptTemplate<\/p>\n<p><span class=\"token comment\"># \u95ee\u9898\u751f\u6210Prompt<\/span><br \/>\nquestion_gen_prompt <span class=\"token operator\">&#061;<\/span> PromptTemplate<span class=\"token punctuation\">(<\/span><br \/>\n    template<span class=\"token operator\">&#061;<\/span><span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u9605\u8bfb\u4ee5\u4e0b\u6587\u6863&#xff0c;\u751f\u62103\u4e2a\u53ef\u80fd\u7684\u95ee\u9898\u3002<\/p>\n<p>\u8981\u6c42:<br \/>\n1. \u95ee\u9898\u5e94\u8be5\u6db5\u76d6\u6587\u6863\u7684\u6838\u5fc3\u5185\u5bb9<br \/>\n2. \u95ee\u9898\u7684\u8868\u8fbe\u65b9\u5f0f\u5e94\u8be5\u591a\u6837\u5316<br \/>\n3. \u6bcf\u4e2a\u95ee\u9898\u72ec\u7acb\u6210\u884c<\/p>\n<p>\u6587\u6863:<br \/>\n{context}<\/p>\n<p>\u95ee\u9898\u5217\u8868:&#034;&#034;&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n    input_variables<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;context&#034;<\/span><span class=\"token punctuation\">]<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token keyword\">def<\/span> <span class=\"token function\">generate_questions<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u4e3a\u6587\u6863\u751f\u6210\u5047\u8bbe\u6027\u95ee\u9898&#034;&#034;&#034;<\/span><br \/>\n    llm <span class=\"token operator\">&#061;<\/span> OpenAI<span class=\"token punctuation\">(<\/span>temperature<span class=\"token operator\">&#061;<\/span><span class=\"token number\">0.7<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    all_questions <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p>    <span class=\"token keyword\">for<\/span> i<span class=\"token punctuation\">,<\/span> doc <span class=\"token keyword\">in<\/span> <span class=\"token builtin\">enumerate<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">,<\/span> <span class=\"token number\">1<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token comment\"># \u751f\u6210\u95ee\u9898<\/span><br \/>\n        response <span class=\"token operator\">&#061;<\/span> llm<span class=\"token punctuation\">(<\/span>question_gen_prompt<span class=\"token punctuation\">.<\/span><span class=\"token builtin\">format<\/span><span class=\"token punctuation\">(<\/span>context<span class=\"token operator\">&#061;<\/span>doc<span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\n        questions <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span>q<span class=\"token punctuation\">.<\/span>strip<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token keyword\">for<\/span> q <span class=\"token keyword\">in<\/span> response<span class=\"token punctuation\">.<\/span>split<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;\\\\n&#039;<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token keyword\">if<\/span> q<span class=\"token punctuation\">.<\/span>strip<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p>        <span class=\"token comment\"># \u5b58\u50a8\u95ee\u9898\u4e0e\u6587\u6863\u7684\u6620\u5c04<\/span><br \/>\n        <span class=\"token keyword\">for<\/span> q <span class=\"token keyword\">in<\/span> questions<span class=\"token punctuation\">:<\/span><br \/>\n            all_questions<span class=\"token punctuation\">.<\/span>append<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">{<\/span><br \/>\n                <span class=\"token string\">&#034;question&#034;<\/span><span class=\"token punctuation\">:<\/span> q<span class=\"token punctuation\">,<\/span><br \/>\n                <span class=\"token string\">&#034;source_doc&#034;<\/span><span class=\"token punctuation\">:<\/span> doc<span class=\"token punctuation\">,<\/span><br \/>\n                <span class=\"token string\">&#034;source_id&#034;<\/span><span class=\"token punctuation\">:<\/span> i<br \/>\n            <span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token keyword\">return<\/span> all_questions<\/p>\n<p><span class=\"token comment\"># \u4f7f\u7528\u793a\u4f8b<\/span><br \/>\nquestions <span class=\"token operator\">&#061;<\/span> generate_questions<span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u7528\u95ee\u9898\u6784\u5efa\u7d22\u5f15<\/span><br \/>\nquestion_docs <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><br \/>\n    Document<span class=\"token punctuation\">(<\/span>page_content<span class=\"token operator\">&#061;<\/span>q<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;question&#034;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> metadata<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;source_id&#034;<\/span><span class=\"token punctuation\">:<\/span> q<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;source_id&#034;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">for<\/span> q <span class=\"token keyword\">in<\/span> questions<br \/>\n<span class=\"token punctuation\">]<\/span><\/p>\n<p>question_index <span class=\"token operator\">&#061;<\/span> Chroma<span class=\"token punctuation\">.<\/span>from_documents<span class=\"token punctuation\">(<\/span>question_docs<span class=\"token punctuation\">,<\/span> embeddings<span class=\"token punctuation\">)<\/span><\/p>\n<h4>3.7 \u591a\u7d22\u5f15\u878d\u5408\u7b56\u7565<\/h4>\n<p>  #mermaid-svg-WzTycUmhToTyUlO9{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-WzTycUmhToTyUlO9 .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-WzTycUmhToTyUlO9 .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-WzTycUmhToTyUlO9 .error-icon{fill:#552222;}#mermaid-svg-WzTycUmhToTyUlO9 .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-WzTycUmhToTyUlO9 .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-WzTycUmhToTyUlO9 .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-WzTycUmhToTyUlO9 .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-WzTycUmhToTyUlO9 .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-WzTycUmhToTyUlO9 .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-WzTycUmhToTyUlO9 .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-WzTycUmhToTyUlO9 .marker{fill:#333333;stroke:#333333;}#mermaid-svg-WzTycUmhToTyUlO9 .marker.cross{stroke:#333333;}#mermaid-svg-WzTycUmhToTyUlO9 svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-WzTycUmhToTyUlO9 p{margin:0;}#mermaid-svg-WzTycUmhToTyUlO9 .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-WzTycUmhToTyUlO9 .cluster-label text{fill:#333;}#mermaid-svg-WzTycUmhToTyUlO9 .cluster-label span{color:#333;}#mermaid-svg-WzTycUmhToTyUlO9 .cluster-label span p{background-color:transparent;}#mermaid-svg-WzTycUmhToTyUlO9 .label text,#mermaid-svg-WzTycUmhToTyUlO9 span{fill:#333;color:#333;}#mermaid-svg-WzTycUmhToTyUlO9 .node rect,#mermaid-svg-WzTycUmhToTyUlO9 .node circle,#mermaid-svg-WzTycUmhToTyUlO9 .node ellipse,#mermaid-svg-WzTycUmhToTyUlO9 .node polygon,#mermaid-svg-WzTycUmhToTyUlO9 .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-WzTycUmhToTyUlO9 .rough-node .label text,#mermaid-svg-WzTycUmhToTyUlO9 .node .label text,#mermaid-svg-WzTycUmhToTyUlO9 .image-shape .label,#mermaid-svg-WzTycUmhToTyUlO9 .icon-shape .label{text-anchor:middle;}#mermaid-svg-WzTycUmhToTyUlO9 .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-WzTycUmhToTyUlO9 .rough-node .label,#mermaid-svg-WzTycUmhToTyUlO9 .node .label,#mermaid-svg-WzTycUmhToTyUlO9 .image-shape .label,#mermaid-svg-WzTycUmhToTyUlO9 .icon-shape .label{text-align:center;}#mermaid-svg-WzTycUmhToTyUlO9 .node.clickable{cursor:pointer;}#mermaid-svg-WzTycUmhToTyUlO9 .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-WzTycUmhToTyUlO9 .arrowheadPath{fill:#333333;}#mermaid-svg-WzTycUmhToTyUlO9 .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-WzTycUmhToTyUlO9 .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-WzTycUmhToTyUlO9 .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-WzTycUmhToTyUlO9 .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-WzTycUmhToTyUlO9 .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-WzTycUmhToTyUlO9 .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-WzTycUmhToTyUlO9 .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-WzTycUmhToTyUlO9 .cluster text{fill:#333;}#mermaid-svg-WzTycUmhToTyUlO9 .cluster span{color:#333;}#mermaid-svg-WzTycUmhToTyUlO9 div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-WzTycUmhToTyUlO9 .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-WzTycUmhToTyUlO9 rect.text{fill:none;stroke-width:0;}#mermaid-svg-WzTycUmhToTyUlO9 .icon-shape,#mermaid-svg-WzTycUmhToTyUlO9 .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-WzTycUmhToTyUlO9 .icon-shape p,#mermaid-svg-WzTycUmhToTyUlO9 .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-WzTycUmhToTyUlO9 .icon-shape rect,#mermaid-svg-WzTycUmhToTyUlO9 .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-WzTycUmhToTyUlO9 .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-WzTycUmhToTyUlO9 .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-WzTycUmhToTyUlO9 :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7528\u6237\u95ee\u9898<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u53e5\u5b50\u7d22\u5f15\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u6458\u8981\u7d22\u5f15\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u95ee\u9898\u7d22\u5f15\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7236\u6587\u6863\u7d22\u5f15\u68c0\u7d22<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u591a\u8def\u53ec\u56de\u5408\u5e76\u7ed3\u679c<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u53bb\u91cd\u79fb\u9664\u91cd\u590d\u6587\u6863<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u91cd\u6392\u5e8fRe-rank<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>Top-N\u7ed3\u679c<\/p>\n<p><\/span> <\/p>\n<p>\u6838\u5fc3\u6982\u5ff5&#xff1a;\u540c\u65f6\u4f7f\u7528\u591a\u79cd\u7d22\u5f15\u7b56\u7565&#xff0c;\u878d\u5408\u6240\u6709\u7ed3\u679c\u3002<\/p>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u591a\u8def\u53ec\u56de&#xff1a;\u4ece\u591a\u4e2a\u7d22\u5f15\u901a\u9053\u540c\u65f6\u68c0\u7d22\u6587\u6863<\/li>\n<li>\u7ed3\u679c\u878d\u5408&#xff1a;\u5c06\u591a\u4e2a\u68c0\u7d22\u7ed3\u679c\u5408\u5e76\u3001\u53bb\u91cd\u3001\u6392\u5e8f<\/li>\n<\/ul>\n<p><span class=\"token keyword\">def<\/span> <span class=\"token function\">multi_index_retrieval<\/span><span class=\"token punctuation\">(<\/span>query<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u591a\u7d22\u5f15\u878d\u5408\u68c0\u7d22&#034;&#034;&#034;<\/span><br \/>\n    all_docs <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<p>    <span class=\"token comment\"># 1. \u53e5\u5b50\u7a97\u53e3\u68c0\u7d22<\/span><br \/>\n    sentence_results <span class=\"token operator\">&#061;<\/span> sentence_retriever<span class=\"token punctuation\">.<\/span>get_relevant_documents<span class=\"token punctuation\">(<\/span>query<span class=\"token punctuation\">)<\/span><br \/>\n    all_docs<span class=\"token punctuation\">.<\/span>extend<span class=\"token punctuation\">(<\/span>sentence_results<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># 2. \u6458\u8981\u68c0\u7d22<\/span><br \/>\n    summary_results <span class=\"token operator\">&#061;<\/span> summary_retriever<span class=\"token punctuation\">.<\/span>get_relevant_documents<span class=\"token punctuation\">(<\/span>query<span class=\"token punctuation\">)<\/span><br \/>\n    all_docs<span class=\"token punctuation\">.<\/span>extend<span class=\"token punctuation\">(<\/span>summary_results<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># 3. \u95ee\u9898\u68c0\u7d22<\/span><br \/>\n    question_results <span class=\"token operator\">&#061;<\/span> question_retriever<span class=\"token punctuation\">.<\/span>get_relevant_documents<span class=\"token punctuation\">(<\/span>query<span class=\"token punctuation\">)<\/span><br \/>\n    all_docs<span class=\"token punctuation\">.<\/span>extend<span class=\"token punctuation\">(<\/span>question_results<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># 4. \u53bb\u91cd<\/span><br \/>\n    unique_docs <span class=\"token operator\">&#061;<\/span> <span class=\"token builtin\">list<\/span><span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">{<\/span>doc<span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">:<\/span> doc <span class=\"token keyword\">for<\/span> doc <span class=\"token keyword\">in<\/span> all_docs<span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">.<\/span>values<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># 5. \u91cd\u6392\u5e8f<\/span><br \/>\n    reranked_docs <span class=\"token operator\">&#061;<\/span> reranker<span class=\"token punctuation\">(<\/span>query<span class=\"token punctuation\">,<\/span> unique_docs<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token keyword\">return<\/span> reranked_docs<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span><span class=\"token number\">5<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<hr \/>\n<h3>\u56db\u3001\u68c0\u7d22\u540e\u589e\u5f3a<\/h3>\n<p>\u68c0\u7d22\u540e\u589e\u5f3a\u7684\u76ee\u6807\u662f\u5bf9\u68c0\u7d22\u7ed3\u679c\u8fdb\u884c\u4f18\u5316&#xff0c;\u63d0\u5347\u6700\u7ec8\u7b54\u6848\u7684\u8d28\u91cf\u3002<\/p>\n<h4>4.1 \u91cd\u6392\u5e8f (Re-ranking) \u539f\u7406<\/h4>\n<p>\u6838\u5fc3\u95ee\u9898&#xff1a;\u5411\u91cf\u68c0\u7d22\u4f7f\u7528\u4f59\u5f26\u76f8\u4f3c\u5ea6&#xff0c;\u4f46\u8fd9\u4e0d\u7b49\u4e8e\u95ee\u9898\u76f8\u5173\u6027&#xff01;<\/p>\n<p>\u4f8b\u5b50<\/p>\n<p>\u7528\u6237\u95ee&#xff1a;\u201c\u5982\u4f55\u51cf\u80a5&#xff1f;\u201d<\/p>\n<ul>\n<li>\u6587\u6863A&#xff08;\u5411\u91cf\u76f8\u4f3c\u5ea60.85&#xff09;&#xff1a;\u201c\u51cf\u80a5\u9700\u8981\u63a7\u5236\u996e\u98df\u548c\u8fd0\u52a8\u201d &#8211; \u76f4\u63a5\u76f8\u5173<\/li>\n<li>\u6587\u6863B&#xff08;\u5411\u91cf\u76f8\u4f3c\u5ea60.87&#xff09;&#xff1a;\u201c\u8fd9\u4e2a\u6587\u6863\u8ba8\u8bba\u4e86\u5404\u79cd\u5065\u5eb7\u95ee\u9898&#xff0c;\u5305\u62ec\u5982\u4f55\u589e\u91cd\u3001\u5982\u4f55\u51cf\u80a5\u2026\u201d &#8211; \u76f8\u5173\u4f46\u5570\u55e6<\/li>\n<\/ul>\n<p>\u91cd\u6392\u5e8f\u540e&#xff1a;\u6587\u6863A\u6392\u7b2c\u4e00&#xff0c;\u56e0\u4e3a\u76f4\u63a5\u56de\u7b54\u95ee\u9898<\/p>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u53cc\u5854\u6a21\u578b&#xff1a;\u5411\u91cf\u68c0\u7d22\u4f7f\u7528\u7684\u6a21\u578b&#xff0c;\u67e5\u8be2\u548c\u6587\u6863\u72ec\u7acb\u7f16\u7801<\/li>\n<li>\u4ea4\u53c9\u7f16\u7801\u5668&#xff1a;\u91cd\u6392\u5e8f\u4f7f\u7528\u7684\u6a21\u578b&#xff0c;\u67e5\u8be2\u548c\u6587\u6863\u4e00\u8d77\u7f16\u7801&#xff0c;\u7cbe\u5ea6\u66f4\u9ad8\u4f46\u901f\u5ea6\u66f4\u6162<\/li>\n<\/ul>\n<p>  #mermaid-svg-6OQAgoxFNffYeUS1{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;fill:#333;}@keyframes edge-animation-frame{from{stroke-dashoffset:0;}}@keyframes dash{to{stroke-dashoffset:0;}}#mermaid-svg-6OQAgoxFNffYeUS1 .edge-animation-slow{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 50s linear infinite;stroke-linecap:round;}#mermaid-svg-6OQAgoxFNffYeUS1 .edge-animation-fast{stroke-dasharray:9,5!important;stroke-dashoffset:900;animation:dash 20s linear infinite;stroke-linecap:round;}#mermaid-svg-6OQAgoxFNffYeUS1 .error-icon{fill:#552222;}#mermaid-svg-6OQAgoxFNffYeUS1 .error-text{fill:#552222;stroke:#552222;}#mermaid-svg-6OQAgoxFNffYeUS1 .edge-thickness-normal{stroke-width:1px;}#mermaid-svg-6OQAgoxFNffYeUS1 .edge-thickness-thick{stroke-width:3.5px;}#mermaid-svg-6OQAgoxFNffYeUS1 .edge-pattern-solid{stroke-dasharray:0;}#mermaid-svg-6OQAgoxFNffYeUS1 .edge-thickness-invisible{stroke-width:0;fill:none;}#mermaid-svg-6OQAgoxFNffYeUS1 .edge-pattern-dashed{stroke-dasharray:3;}#mermaid-svg-6OQAgoxFNffYeUS1 .edge-pattern-dotted{stroke-dasharray:2;}#mermaid-svg-6OQAgoxFNffYeUS1 .marker{fill:#333333;stroke:#333333;}#mermaid-svg-6OQAgoxFNffYeUS1 .marker.cross{stroke:#333333;}#mermaid-svg-6OQAgoxFNffYeUS1 svg{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:16px;}#mermaid-svg-6OQAgoxFNffYeUS1 p{margin:0;}#mermaid-svg-6OQAgoxFNffYeUS1 .label{font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;color:#333;}#mermaid-svg-6OQAgoxFNffYeUS1 .cluster-label text{fill:#333;}#mermaid-svg-6OQAgoxFNffYeUS1 .cluster-label span{color:#333;}#mermaid-svg-6OQAgoxFNffYeUS1 .cluster-label span p{background-color:transparent;}#mermaid-svg-6OQAgoxFNffYeUS1 .label text,#mermaid-svg-6OQAgoxFNffYeUS1 span{fill:#333;color:#333;}#mermaid-svg-6OQAgoxFNffYeUS1 .node rect,#mermaid-svg-6OQAgoxFNffYeUS1 .node circle,#mermaid-svg-6OQAgoxFNffYeUS1 .node ellipse,#mermaid-svg-6OQAgoxFNffYeUS1 .node polygon,#mermaid-svg-6OQAgoxFNffYeUS1 .node path{fill:#ECECFF;stroke:#9370DB;stroke-width:1px;}#mermaid-svg-6OQAgoxFNffYeUS1 .rough-node .label text,#mermaid-svg-6OQAgoxFNffYeUS1 .node .label text,#mermaid-svg-6OQAgoxFNffYeUS1 .image-shape .label,#mermaid-svg-6OQAgoxFNffYeUS1 .icon-shape .label{text-anchor:middle;}#mermaid-svg-6OQAgoxFNffYeUS1 .node .katex path{fill:#000;stroke:#000;stroke-width:1px;}#mermaid-svg-6OQAgoxFNffYeUS1 .rough-node .label,#mermaid-svg-6OQAgoxFNffYeUS1 .node .label,#mermaid-svg-6OQAgoxFNffYeUS1 .image-shape .label,#mermaid-svg-6OQAgoxFNffYeUS1 .icon-shape .label{text-align:center;}#mermaid-svg-6OQAgoxFNffYeUS1 .node.clickable{cursor:pointer;}#mermaid-svg-6OQAgoxFNffYeUS1 .root .anchor path{fill:#333333!important;stroke-width:0;stroke:#333333;}#mermaid-svg-6OQAgoxFNffYeUS1 .arrowheadPath{fill:#333333;}#mermaid-svg-6OQAgoxFNffYeUS1 .edgePath .path{stroke:#333333;stroke-width:2.0px;}#mermaid-svg-6OQAgoxFNffYeUS1 .flowchart-link{stroke:#333333;fill:none;}#mermaid-svg-6OQAgoxFNffYeUS1 .edgeLabel{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-6OQAgoxFNffYeUS1 .edgeLabel p{background-color:rgba(232,232,232, 0.8);}#mermaid-svg-6OQAgoxFNffYeUS1 .edgeLabel rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-6OQAgoxFNffYeUS1 .labelBkg{background-color:rgba(232, 232, 232, 0.5);}#mermaid-svg-6OQAgoxFNffYeUS1 .cluster rect{fill:#ffffde;stroke:#aaaa33;stroke-width:1px;}#mermaid-svg-6OQAgoxFNffYeUS1 .cluster text{fill:#333;}#mermaid-svg-6OQAgoxFNffYeUS1 .cluster span{color:#333;}#mermaid-svg-6OQAgoxFNffYeUS1 div.mermaidTooltip{position:absolute;text-align:center;max-width:200px;padding:2px;font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;font-size:12px;background:hsl(80, 100%, 96.2745098039%);border:1px solid #aaaa33;border-radius:2px;pointer-events:none;z-index:100;}#mermaid-svg-6OQAgoxFNffYeUS1 .flowchartTitleText{text-anchor:middle;font-size:18px;fill:#333;}#mermaid-svg-6OQAgoxFNffYeUS1 rect.text{fill:none;stroke-width:0;}#mermaid-svg-6OQAgoxFNffYeUS1 .icon-shape,#mermaid-svg-6OQAgoxFNffYeUS1 .image-shape{background-color:rgba(232,232,232, 0.8);text-align:center;}#mermaid-svg-6OQAgoxFNffYeUS1 .icon-shape p,#mermaid-svg-6OQAgoxFNffYeUS1 .image-shape p{background-color:rgba(232,232,232, 0.8);padding:2px;}#mermaid-svg-6OQAgoxFNffYeUS1 .icon-shape rect,#mermaid-svg-6OQAgoxFNffYeUS1 .image-shape rect{opacity:0.5;background-color:rgba(232,232,232, 0.8);fill:rgba(232,232,232, 0.8);}#mermaid-svg-6OQAgoxFNffYeUS1 .label-icon{display:inline-block;height:1em;overflow:visible;vertical-align:-0.125em;}#mermaid-svg-6OQAgoxFNffYeUS1 .node .label-icon path{fill:currentColor;stroke:revert;stroke-width:revert;}#mermaid-svg-6OQAgoxFNffYeUS1 :root{&#8211;mermaid-font-family:\\&#8221;trebuchet ms\\&#8221;,verdana,arial,sans-serif;} <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"edgeLabel\"><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u5411\u91cf\u68c0\u7d22Top-10\u6587\u6863<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>\u7c97\u6392\u7ed3\u679c<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>Re-rank\u6a21\u578b\u7cbe\u6253\u7ec6\u7b97<\/p>\n<p><\/span> <\/p>\n<p>         <span class=\"nodeLabel\"><\/p>\n<p>Top-3\u9ad8\u8d28\u91cf\u6587\u6863<\/p>\n<p><\/span> <\/p>\n<h4>4.2 \u4e3b\u6d41Re-ranker\u5bf9\u6bd4<\/h4>\n<table>\n<tr>\u6a21\u578b\u7c7b\u578b\u7279\u70b9\u901f\u5ea6\u51c6\u786e\u7387<\/tr>\n<tbody>\n<tr>\n<td>Cohere Rerank-v3.5<\/td>\n<td>\u5546\u4e1aAPI<\/td>\n<td>\u591a\u8bed\u8a00&#xff0c;\u6548\u679c\u6700\u597d<\/td>\n<td>\u4e2d\u7b49<\/td>\n<td>\u6700\u9ad8<\/td>\n<\/tr>\n<tr>\n<td>BGE Reranker-v2<\/td>\n<td>\u5f00\u6e90<\/td>\n<td>\u4e2d\u6587\u53cb\u597d&#xff0c;\u514d\u8d39<\/td>\n<td>\u8f83\u5feb<\/td>\n<td>\u9ad8<\/td>\n<\/tr>\n<tr>\n<td>FlashRank<\/td>\n<td>\u5f00\u6e90<\/td>\n<td>\u6781\u5feb&#xff0c;\u8f7b\u91cf\u7ea7<\/td>\n<td>\u6781\u5feb<\/td>\n<td>\u4e2d\u7b49<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h4>4.3 Re-ranking\u5b9e\u73b0<\/h4>\n<p><span class=\"token comment\"># \u65b9\u68481: Cohere Rerank<\/span><br \/>\n<span class=\"token keyword\">import<\/span> cohere<\/p>\n<p><span class=\"token keyword\">def<\/span> <span class=\"token function\">cohere_rerank<\/span><span class=\"token punctuation\">(<\/span>query<span class=\"token punctuation\">,<\/span> documents<span class=\"token punctuation\">,<\/span> top_n<span class=\"token operator\">&#061;<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    co <span class=\"token operator\">&#061;<\/span> cohere<span class=\"token punctuation\">.<\/span>Client<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;your-api-key&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    results <span class=\"token operator\">&#061;<\/span> co<span class=\"token punctuation\">.<\/span>rerank<span class=\"token punctuation\">(<\/span><br \/>\n        model<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;rerank-v3.5&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        query<span class=\"token operator\">&#061;<\/span>query<span class=\"token punctuation\">,<\/span><br \/>\n        documents<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span>doc<span class=\"token punctuation\">.<\/span>page_content <span class=\"token keyword\">for<\/span> doc <span class=\"token keyword\">in<\/span> documents<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        top_n<span class=\"token operator\">&#061;<\/span>top_n<br \/>\n    <span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">return<\/span> <span class=\"token punctuation\">[<\/span>documents<span class=\"token punctuation\">[<\/span>r<span class=\"token punctuation\">.<\/span>index<span class=\"token punctuation\">]<\/span> <span class=\"token keyword\">for<\/span> r <span class=\"token keyword\">in<\/span> results<span class=\"token punctuation\">.<\/span>results<span class=\"token punctuation\">]<\/span><\/p>\n<p><span class=\"token comment\"># \u65b9\u68482: BGE Reranker<\/span><br \/>\n<span class=\"token keyword\">from<\/span> sentence_transformers <span class=\"token keyword\">import<\/span> CrossEncoder<\/p>\n<p><span class=\"token keyword\">def<\/span> <span class=\"token function\">bge_rerank<\/span><span class=\"token punctuation\">(<\/span>query<span class=\"token punctuation\">,<\/span> documents<span class=\"token punctuation\">,<\/span> top_k<span class=\"token operator\">&#061;<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    model <span class=\"token operator\">&#061;<\/span> CrossEncoder<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;BAAI\/bge-reranker-v2-m3&#039;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u6784\u5efa\u67e5\u8be2-\u6587\u6863\u5bf9<\/span><br \/>\n    pairs <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">[<\/span>query<span class=\"token punctuation\">,<\/span> doc<span class=\"token punctuation\">.<\/span>page_content<span class=\"token punctuation\">]<\/span> <span class=\"token keyword\">for<\/span> doc <span class=\"token keyword\">in<\/span> documents<span class=\"token punctuation\">]<\/span><\/p>\n<p>    <span class=\"token comment\"># \u8ba1\u7b97\u5206\u6570<\/span><br \/>\n    scores <span class=\"token operator\">&#061;<\/span> model<span class=\"token punctuation\">.<\/span>predict<span class=\"token punctuation\">(<\/span>pairs<span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token comment\"># \u6392\u5e8f<\/span><br \/>\n    scored_docs <span class=\"token operator\">&#061;<\/span> <span class=\"token builtin\">sorted<\/span><span class=\"token punctuation\">(<\/span><span class=\"token builtin\">zip<\/span><span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">,<\/span> scores<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span> key<span class=\"token operator\">&#061;<\/span><span class=\"token keyword\">lambda<\/span> x<span class=\"token punctuation\">:<\/span> x<span class=\"token punctuation\">[<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> reverse<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    <span class=\"token keyword\">return<\/span> <span class=\"token punctuation\">[<\/span>doc <span class=\"token keyword\">for<\/span> doc<span class=\"token punctuation\">,<\/span> score <span class=\"token keyword\">in<\/span> scored_docs<span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">:<\/span>top_k<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">]<\/span><\/p>\n<h4>4.4 \u4e0a\u4e0b\u6587\u538b\u7f29<\/h4>\n<p>\u6838\u5fc3\u6982\u5ff5&#xff1a;\u5bf9\u68c0\u7d22\u5230\u7684\u6587\u6863\u8fdb\u884c\u538b\u7f29&#xff0c;\u53bb\u9664\u65e0\u5173\u5185\u5bb9&#xff0c;\u4fdd\u7559\u5173\u952e\u4fe1\u606f\u3002<\/p>\n<p>\u4e13\u4e1a\u672f\u8bed\u89e3\u91ca&#xff1a;<\/p>\n<ul>\n<li>\u4e0a\u4e0b\u6587\u538b\u7f29&#xff1a;\u63d0\u53d6\u6587\u6863\u4e2d\u4e0e\u95ee\u9898\u6700\u76f8\u5173\u7684\u7247\u6bb5<\/li>\n<li>\u4fe1\u606f\u5bc6\u5ea6&#xff1a;\u5355\u4f4d\u957f\u5ea6\u5185\u7684\u6709\u6548\u4fe1\u606f\u91cf<\/li>\n<\/ul>\n<p><span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>retrievers <span class=\"token keyword\">import<\/span> ContextualCompressionRetriever<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>document_compressors <span class=\"token keyword\">import<\/span> LLMChainExtractor<\/p>\n<p><span class=\"token comment\"># \u521b\u5efa\u538b\u7f29\u5668<\/span><br \/>\ncompressor <span class=\"token operator\">&#061;<\/span> LLMChainExtractor<span class=\"token punctuation\">.<\/span>from_llm<span class=\"token punctuation\">(<\/span>ChatOpenAI<span class=\"token punctuation\">(<\/span>model<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;gpt-4&#034;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u521b\u5efa\u538b\u7f29\u68c0\u7d22\u5668<\/span><br \/>\ncompression_retriever <span class=\"token operator\">&#061;<\/span> ContextualCompressionRetriever<span class=\"token punctuation\">(<\/span><br \/>\n    base_compressor<span class=\"token operator\">&#061;<\/span>compressor<span class=\"token punctuation\">,<\/span><br \/>\n    base_retriever<span class=\"token operator\">&#061;<\/span>vectorstore<span class=\"token punctuation\">.<\/span>as_retriever<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token punctuation\">)<\/span><\/p>\n<p><span class=\"token comment\"># \u68c0\u7d22\u5e76\u538b\u7f29<\/span><br \/>\ncompressed_docs <span class=\"token operator\">&#061;<\/span> compression_retriever<span class=\"token punctuation\">.<\/span>get_relevant_documents<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#034;\u4ec0\u4e48\u662fRAG&#xff1f;&#034;<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<hr \/>\n<h3>\u4e94\u3001\u5b8c\u6574\u751f\u4ea7\u7ea7\u793a\u4f8b<\/h3>\n<h4>5.1 LangChain\u751f\u4ea7\u7ea7\u5b9e\u73b0<\/h4>\n<p><span class=\"token keyword\">import<\/span> os<br \/>\n<span class=\"token keyword\">from<\/span> typing <span class=\"token keyword\">import<\/span> List<span class=\"token punctuation\">,<\/span> Dict<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>embeddings <span class=\"token keyword\">import<\/span> OpenAIEmbeddings<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>vectorstores <span class=\"token keyword\">import<\/span> Chroma<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>chat_models <span class=\"token keyword\">import<\/span> ChatOpenAI<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>chains <span class=\"token keyword\">import<\/span> RetrievalQA<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>retrievers <span class=\"token keyword\">import<\/span> ContextualCompressionRetriever<br \/>\n<span class=\"token keyword\">from<\/span> langchain_community<span class=\"token punctuation\">.<\/span>document_compressors<span class=\"token punctuation\">.<\/span>cohere_rerank <span class=\"token keyword\">import<\/span> CohereRerank<br \/>\n<span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>retrievers <span class=\"token keyword\">import<\/span> EnsembleRetriever<br \/>\n<span class=\"token keyword\">from<\/span> langchain_community<span class=\"token punctuation\">.<\/span>retrievers <span class=\"token keyword\">import<\/span> BM25Retriever<\/p>\n<p><span class=\"token keyword\">class<\/span> <span class=\"token class-name\">ProductionRAG<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u751f\u4ea7\u7ea7RAG\u7cfb\u7edf&#034;&#034;&#034;<\/span><\/p>\n<p>    <span class=\"token keyword\">def<\/span> <span class=\"token function\">__init__<\/span><span class=\"token punctuation\">(<\/span><br \/>\n        self<span class=\"token punctuation\">,<\/span><br \/>\n        knowledge_base<span class=\"token punctuation\">:<\/span> List<span class=\"token punctuation\">[<\/span><span class=\"token builtin\">str<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        model<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token string\">&#034;gpt-4&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        top_k<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">int<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token number\">20<\/span><span class=\"token punctuation\">,<\/span><br \/>\n        rerank_top_n<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">int<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token number\">5<\/span><br \/>\n    <span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token comment\"># \u521d\u59cb\u5316\u5d4c\u5165\u6a21\u578b<\/span><br \/>\n        self<span class=\"token punctuation\">.<\/span>embeddings <span class=\"token operator\">&#061;<\/span> OpenAIEmbeddings<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u521b\u5efa\u5411\u91cf\u6570\u636e\u5e93<\/span><br \/>\n        self<span class=\"token punctuation\">.<\/span>vectorstore <span class=\"token operator\">&#061;<\/span> Chroma<span class=\"token punctuation\">(<\/span><br \/>\n            collection_name<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;production_kb&#034;<\/span><span class=\"token punctuation\">,<\/span><br \/>\n            embedding_function<span class=\"token operator\">&#061;<\/span>self<span class=\"token punctuation\">.<\/span>embeddings<br \/>\n        <span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u6dfb\u52a0\u6587\u6863<\/span><br \/>\n        <span class=\"token keyword\">if<\/span> self<span class=\"token punctuation\">.<\/span>vectorstore<span class=\"token punctuation\">.<\/span>_collection<span class=\"token punctuation\">.<\/span>count<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&#061;&#061;<\/span> <span class=\"token number\">0<\/span><span class=\"token punctuation\">:<\/span><br \/>\n            <span class=\"token keyword\">from<\/span> langchain<span class=\"token punctuation\">.<\/span>schema <span class=\"token keyword\">import<\/span> Document<br \/>\n            docs <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">[<\/span>Document<span class=\"token punctuation\">(<\/span>page_content<span class=\"token operator\">&#061;<\/span>text<span class=\"token punctuation\">)<\/span> <span class=\"token keyword\">for<\/span> text <span class=\"token keyword\">in<\/span> knowledge_base<span class=\"token punctuation\">]<\/span><br \/>\n            self<span class=\"token punctuation\">.<\/span>vectorstore<span class=\"token punctuation\">.<\/span>add_documents<span class=\"token punctuation\">(<\/span>docs<span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u6df7\u5408\u68c0\u7d22\u5668<\/span><br \/>\n        bm25_retriever <span class=\"token operator\">&#061;<\/span> BM25Retriever<span class=\"token punctuation\">.<\/span>from_documents<span class=\"token punctuation\">(<\/span>docs<span class=\"token punctuation\">,<\/span> k<span class=\"token operator\">&#061;<\/span>top_k<span class=\"token punctuation\">)<\/span><br \/>\n        vector_retriever <span class=\"token operator\">&#061;<\/span> self<span class=\"token punctuation\">.<\/span>vectorstore<span class=\"token punctuation\">.<\/span>as_retriever<span class=\"token punctuation\">(<\/span>search_kwargs<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;k&#034;<\/span><span class=\"token punctuation\">:<\/span> top_k<span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u878d\u5408\u68c0\u7d22\u5668<\/span><br \/>\n        self<span class=\"token punctuation\">.<\/span>retriever <span class=\"token operator\">&#061;<\/span> EnsembleRetriever<span class=\"token punctuation\">(<\/span><br \/>\n            retrievers<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span>bm25_retriever<span class=\"token punctuation\">,<\/span> vector_retriever<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n            weights<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token number\">0.4<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">0.6<\/span><span class=\"token punctuation\">]<\/span><br \/>\n        <span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u91cd\u6392\u5e8f<\/span><br \/>\n        compressor <span class=\"token operator\">&#061;<\/span> CohereRerank<span class=\"token punctuation\">(<\/span>top_n<span class=\"token operator\">&#061;<\/span>rerank_top_n<span class=\"token punctuation\">)<\/span><br \/>\n        self<span class=\"token punctuation\">.<\/span>compression_retriever <span class=\"token operator\">&#061;<\/span> ContextualCompressionRetriever<span class=\"token punctuation\">(<\/span><br \/>\n            base_compressor<span class=\"token operator\">&#061;<\/span>compressor<span class=\"token punctuation\">,<\/span><br \/>\n            base_retriever<span class=\"token operator\">&#061;<\/span>self<span class=\"token punctuation\">.<\/span>retriever<br \/>\n        <span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u521b\u5efaRAG\u94fe<\/span><br \/>\n        self<span class=\"token punctuation\">.<\/span>qa_chain <span class=\"token operator\">&#061;<\/span> RetrievalQA<span class=\"token punctuation\">.<\/span>from_chain_type<span class=\"token punctuation\">(<\/span><br \/>\n            llm<span class=\"token operator\">&#061;<\/span>ChatOpenAI<span class=\"token punctuation\">(<\/span>model<span class=\"token operator\">&#061;<\/span>model<span class=\"token punctuation\">,<\/span> temperature<span class=\"token operator\">&#061;<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span><br \/>\n            retriever<span class=\"token operator\">&#061;<\/span>self<span class=\"token punctuation\">.<\/span>compression_retriever<span class=\"token punctuation\">,<\/span><br \/>\n            return_source_documents<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span><br \/>\n        <span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token keyword\">def<\/span> <span class=\"token function\">query<\/span><span class=\"token punctuation\">(<\/span>self<span class=\"token punctuation\">,<\/span> question<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&#8211;<\/span><span class=\"token operator\">&gt;<\/span> Dict<span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u6267\u884c\u67e5\u8be2&#034;&#034;&#034;<\/span><br \/>\n        result <span class=\"token operator\">&#061;<\/span> self<span class=\"token punctuation\">.<\/span>qa_chain<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">{<\/span><span class=\"token string\">&#034;query&#034;<\/span><span class=\"token punctuation\">:<\/span> question<span class=\"token punctuation\">}<\/span><span class=\"token punctuation\">)<\/span><br \/>\n        <span class=\"token keyword\">return<\/span> <span class=\"token punctuation\">{<\/span><br \/>\n            <span class=\"token string\">&#034;question&#034;<\/span><span class=\"token punctuation\">:<\/span> question<span class=\"token punctuation\">,<\/span><br \/>\n            <span class=\"token string\">&#034;answer&#034;<\/span><span class=\"token punctuation\">:<\/span> result<span class=\"token punctuation\">[<\/span><span class=\"token string\">&#034;result&#034;<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n            <span class=\"token string\">&#034;sources&#034;<\/span><span class=\"token punctuation\">:<\/span> result<span class=\"token punctuation\">.<\/span>get<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#034;source_documents&#034;<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token punctuation\">[<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span><br \/>\n        <span class=\"token punctuation\">}<\/span><\/p>\n<h4>5.2 LlamaIndex\u751f\u4ea7\u7ea7\u5b9e\u73b0<\/h4>\n<p><span class=\"token keyword\">from<\/span> llama_index <span class=\"token keyword\">import<\/span> VectorStoreIndex<span class=\"token punctuation\">,<\/span> ServiceContext<br \/>\n<span class=\"token keyword\">from<\/span> llama_index<span class=\"token punctuation\">.<\/span>llms <span class=\"token keyword\">import<\/span> OpenAI<br \/>\n<span class=\"token keyword\">from<\/span> llama_index<span class=\"token punctuation\">.<\/span>embeddings <span class=\"token keyword\">import<\/span> OpenAIEmbedding<br \/>\n<span class=\"token keyword\">from<\/span> llama_index<span class=\"token punctuation\">.<\/span>node_parser <span class=\"token keyword\">import<\/span> SentenceWindowNodeParser<br \/>\n<span class=\"token keyword\">from<\/span> llama_index<span class=\"token punctuation\">.<\/span>postprocessor <span class=\"token keyword\">import<\/span> MetadataReplacementPostprocessor<\/p>\n<p><span class=\"token keyword\">class<\/span> <span class=\"token class-name\">ProductionRAGLlama<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    <span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u751f\u4ea7\u7ea7RAG\u7cfb\u7edf &#8211; LlamaIndex\u7248\u672c&#034;&#034;&#034;<\/span><\/p>\n<p>    <span class=\"token keyword\">def<\/span> <span class=\"token function\">__init__<\/span><span class=\"token punctuation\">(<\/span>self<span class=\"token punctuation\">,<\/span> data_dir<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token string\">&#034;.\/data&#034;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token comment\"># \u914d\u7f6e\u670d\u52a1\u4e0a\u4e0b\u6587<\/span><br \/>\n        self<span class=\"token punctuation\">.<\/span>service_context <span class=\"token operator\">&#061;<\/span> ServiceContext<span class=\"token punctuation\">.<\/span>from_defaults<span class=\"token punctuation\">(<\/span><br \/>\n            llm<span class=\"token operator\">&#061;<\/span>OpenAI<span class=\"token punctuation\">(<\/span>model<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;gpt-4&#034;<\/span><span class=\"token punctuation\">,<\/span> temperature<span class=\"token operator\">&#061;<\/span><span class=\"token number\">0<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span><br \/>\n            embed_model<span class=\"token operator\">&#061;<\/span>OpenAIEmbedding<span class=\"token punctuation\">(<\/span>model<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;text-embedding-ada-002&#034;<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span><br \/>\n            chunk_size<span class=\"token operator\">&#061;<\/span><span class=\"token number\">512<\/span><span class=\"token punctuation\">,<\/span><br \/>\n            chunk_overlap<span class=\"token operator\">&#061;<\/span><span class=\"token number\">50<\/span><br \/>\n        <span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u52a0\u8f7d\u6587\u6863<\/span><br \/>\n        <span class=\"token keyword\">from<\/span> llama_index <span class=\"token keyword\">import<\/span> SimpleDirectoryReader<br \/>\n        documents <span class=\"token operator\">&#061;<\/span> SimpleDirectoryReader<span class=\"token punctuation\">(<\/span>data_dir<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">.<\/span>load_data<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u53e5\u5b50\u7a97\u53e3\u89e3\u6790\u5668<\/span><br \/>\n        node_parser <span class=\"token operator\">&#061;<\/span> SentenceWindowNodeParser<span class=\"token punctuation\">.<\/span>from_defaults<span class=\"token punctuation\">(<\/span>window_size<span class=\"token operator\">&#061;<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">)<\/span><br \/>\n        nodes <span class=\"token operator\">&#061;<\/span> node_parser<span class=\"token punctuation\">.<\/span>get_nodes_from_documents<span class=\"token punctuation\">(<\/span>documents<span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u521b\u5efa\u7d22\u5f15<\/span><br \/>\n        self<span class=\"token punctuation\">.<\/span>index <span class=\"token operator\">&#061;<\/span> VectorStoreIndex<span class=\"token punctuation\">(<\/span>nodes<span class=\"token punctuation\">,<\/span> service_context<span class=\"token operator\">&#061;<\/span>self<span class=\"token punctuation\">.<\/span>service_context<span class=\"token punctuation\">)<\/span><\/p>\n<p>        <span class=\"token comment\"># \u521b\u5efa\u67e5\u8be2\u5f15\u64ce<\/span><br \/>\n        self<span class=\"token punctuation\">.<\/span>query_engine <span class=\"token operator\">&#061;<\/span> self<span class=\"token punctuation\">.<\/span>index<span class=\"token punctuation\">.<\/span>as_query_engine<span class=\"token punctuation\">(<\/span><br \/>\n            similarity_top_k<span class=\"token operator\">&#061;<\/span><span class=\"token number\">10<\/span><span class=\"token punctuation\">,<\/span><br \/>\n            node_postprocessors<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><br \/>\n                MetadataReplacementPostprocessor<span class=\"token punctuation\">(<\/span>target_metadata_key<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;window&#034;<\/span><span class=\"token punctuation\">)<\/span><br \/>\n            <span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span><br \/>\n            response_mode<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#034;compact&#034;<\/span><br \/>\n        <span class=\"token punctuation\">)<\/span><\/p>\n<p>    <span class=\"token keyword\">def<\/span> <span class=\"token function\">query<\/span><span class=\"token punctuation\">(<\/span>self<span class=\"token punctuation\">,<\/span> question<span class=\"token punctuation\">:<\/span> <span class=\"token builtin\">str<\/span><span class=\"token punctuation\">)<\/span> <span class=\"token operator\">&#8211;<\/span><span class=\"token operator\">&gt;<\/span> <span class=\"token builtin\">str<\/span><span class=\"token punctuation\">:<\/span><br \/>\n        <span class=\"token triple-quoted-string string\">&#034;&#034;&#034;\u6267\u884c\u67e5\u8be2&#034;&#034;&#034;<\/span><br \/>\n        response <span class=\"token operator\">&#061;<\/span> self<span class=\"token punctuation\">.<\/span>query_engine<span class=\"token punctuation\">.<\/span>query<span class=\"token punctuation\">(<\/span>question<span class=\"token punctuation\">)<\/span><br \/>\n        <span class=\"token keyword\">return<\/span> <span class=\"token builtin\">str<\/span><span class=\"token punctuation\">(<\/span>response<span class=\"token punctuation\">)<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u76ee\u5f55 RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357\u4e00\u3001\u4e09\u4ee3RAG\u67b6\u6784\u5bf9\u6bd41.1 \u67b6\u6784\u6f14\u8fdb\u65f6\u95f4\u7ebf1.2 \u4e09\u4ee3\u67b6\u6784\u8be6\u7ec6\u5bf9\u6bd41.3 Naive RAG &#8211; \u57fa\u7840\u67b6\u6784Python\u5b9e\u73b0\u793a\u4f8b 1.4 Advanced RAG &#8211; \u589e\u5f3a\u67b6\u67841.5 Modular RAG &#8211; \u6a21\u5757\u5316\u67b6\u67841.6 RAG \u5168\u94fe\u8def\u589e\u5f3a\u8def\u7ebf \u4e8c\u3001\u68c0\u7d22\u524d\u589e\u5f3a2.1 HyDE (Hypothetical Document Embeddings)2.2 Query2Doc2.3 \u67e5\u8be2\u91cd\u5199 (Query Rewriting)2.4 \u67e5\u8be2\u5206\u89e3 (<\/p>\n","protected":false},"author":2,"featured_media":68850,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[52,997,347,2946,81],"topic":[],"class_list":["post-68851","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-server","tag-aigc","tag-langchain","tag-llama","tag-prompt","tag-python"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v20.3 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.wsisp.com\/helps\/68851.html\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"og:description\" content=\"\u76ee\u5f55 RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357\u4e00\u3001\u4e09\u4ee3RAG\u67b6\u6784\u5bf9\u6bd41.1 \u67b6\u6784\u6f14\u8fdb\u65f6\u95f4\u7ebf1.2 \u4e09\u4ee3\u67b6\u6784\u8be6\u7ec6\u5bf9\u6bd41.3 Naive RAG - \u57fa\u7840\u67b6\u6784Python\u5b9e\u73b0\u793a\u4f8b 1.4 Advanced RAG - \u589e\u5f3a\u67b6\u67841.5 Modular RAG - \u6a21\u5757\u5316\u67b6\u67841.6 RAG \u5168\u94fe\u8def\u589e\u5f3a\u8def\u7ebf \u4e8c\u3001\u68c0\u7d22\u524d\u589e\u5f3a2.1 HyDE (Hypothetical Document Embeddings)2.2 Query2Doc2.3 \u67e5\u8be2\u91cd\u5199 (Query Rewriting)2.4 \u67e5\u8be2\u5206\u89e3 (\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.wsisp.com\/helps\/68851.html\" \/>\n<meta property=\"og:site_name\" content=\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"article:published_time\" content=\"2026-01-30T11:49:01+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260130114859-697c9aabec580.png\" \/>\n<meta name=\"author\" content=\"admin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"43 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/68851.html\",\"url\":\"https:\/\/www.wsisp.com\/helps\/68851.html\",\"name\":\"RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"isPartOf\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\"},\"datePublished\":\"2026-01-30T11:49:01+00:00\",\"dateModified\":\"2026-01-30T11:49:01+00:00\",\"author\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/68851.html#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.wsisp.com\/helps\/68851.html\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/68851.html#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.wsisp.com\/helps\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\",\"url\":\"https:\/\/www.wsisp.com\/helps\/\",\"name\":\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"description\":\"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\",\"name\":\"admin\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"contentUrl\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"caption\":\"admin\"},\"sameAs\":[\"http:\/\/wp.wsisp.com\"],\"url\":\"https:\/\/www.wsisp.com\/helps\/author\/admin\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.wsisp.com\/helps\/68851.html","og_locale":"zh_CN","og_type":"article","og_title":"RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","og_description":"\u76ee\u5f55 RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357\u4e00\u3001\u4e09\u4ee3RAG\u67b6\u6784\u5bf9\u6bd41.1 \u67b6\u6784\u6f14\u8fdb\u65f6\u95f4\u7ebf1.2 \u4e09\u4ee3\u67b6\u6784\u8be6\u7ec6\u5bf9\u6bd41.3 Naive RAG - \u57fa\u7840\u67b6\u6784Python\u5b9e\u73b0\u793a\u4f8b 1.4 Advanced RAG - \u589e\u5f3a\u67b6\u67841.5 Modular RAG - \u6a21\u5757\u5316\u67b6\u67841.6 RAG \u5168\u94fe\u8def\u589e\u5f3a\u8def\u7ebf \u4e8c\u3001\u68c0\u7d22\u524d\u589e\u5f3a2.1 HyDE (Hypothetical Document Embeddings)2.2 Query2Doc2.3 \u67e5\u8be2\u91cd\u5199 (Query Rewriting)2.4 \u67e5\u8be2\u5206\u89e3 (","og_url":"https:\/\/www.wsisp.com\/helps\/68851.html","og_site_name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","article_published_time":"2026-01-30T11:49:01+00:00","og_image":[{"url":"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/01\/20260130114859-697c9aabec580.png"}],"author":"admin","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"admin","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"43 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.wsisp.com\/helps\/68851.html","url":"https:\/\/www.wsisp.com\/helps\/68851.html","name":"RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","isPartOf":{"@id":"https:\/\/www.wsisp.com\/helps\/#website"},"datePublished":"2026-01-30T11:49:01+00:00","dateModified":"2026-01-30T11:49:01+00:00","author":{"@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41"},"breadcrumb":{"@id":"https:\/\/www.wsisp.com\/helps\/68851.html#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.wsisp.com\/helps\/68851.html"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.wsisp.com\/helps\/68851.html#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.wsisp.com\/helps"},{"@type":"ListItem","position":2,"name":"RAG\u68c0\u7d22\u589e\u5f3a\u4f18\u5316\u6307\u5357"}]},{"@type":"WebSite","@id":"https:\/\/www.wsisp.com\/helps\/#website","url":"https:\/\/www.wsisp.com\/helps\/","name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","description":"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41","name":"admin","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/","url":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","contentUrl":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","caption":"admin"},"sameAs":["http:\/\/wp.wsisp.com"],"url":"https:\/\/www.wsisp.com\/helps\/author\/admin"}]}},"_links":{"self":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/68851","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/comments?post=68851"}],"version-history":[{"count":0,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/68851\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media\/68850"}],"wp:attachment":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media?parent=68851"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/categories?post=68851"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/tags?post=68851"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/topic?post=68851"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}